US20010049770A1 - Buffer memory management in a system having multiple execution entities - Google Patents
Buffer memory management in a system having multiple execution entities Download PDFInfo
- Publication number
- US20010049770A1 US20010049770A1 US09/207,278 US20727898A US2001049770A1 US 20010049770 A1 US20010049770 A1 US 20010049770A1 US 20727898 A US20727898 A US 20727898A US 2001049770 A1 US2001049770 A1 US 2001049770A1
- Authority
- US
- United States
- Prior art keywords
- cache
- memory
- execution
- instruction
- execution entities
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 230000015654 memory Effects 0.000 title claims abstract description 158
- 239000000872 buffer Substances 0.000 title claims description 39
- 238000000034 method Methods 0.000 claims abstract description 26
- 230000008569 process Effects 0.000 claims abstract description 18
- 230000008520 organization Effects 0.000 claims description 5
- 230000002123 temporal effect Effects 0.000 claims description 3
- 238000013507 mapping Methods 0.000 description 7
- 238000010586 diagram Methods 0.000 description 5
- 238000012986 modification Methods 0.000 description 5
- 230000004048 modification Effects 0.000 description 5
- 238000012546 transfer Methods 0.000 description 5
- 230000008901 benefit Effects 0.000 description 4
- 230000002093 peripheral effect Effects 0.000 description 4
- 238000012545 processing Methods 0.000 description 4
- 230000006870 function Effects 0.000 description 3
- 230000003068 static effect Effects 0.000 description 2
- 238000003491 array Methods 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 230000006835 compression Effects 0.000 description 1
- 238000007906 compression Methods 0.000 description 1
- 230000008878 coupling Effects 0.000 description 1
- 238000010168 coupling process Methods 0.000 description 1
- 238000005859 coupling reaction Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 230000007334 memory performance Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 238000013519 translation Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F12/00—Accessing, addressing or allocating within memory systems or architectures
- G06F12/02—Addressing or allocation; Relocation
- G06F12/08—Addressing or allocation; Relocation in hierarchically structured memory systems, e.g. virtual memory systems
- G06F12/0802—Addressing of a memory level in which the access to the desired data or data block requires associative addressing means, e.g. caches
- G06F12/0844—Multiple simultaneous or quasi-simultaneous cache accessing
- G06F12/0846—Cache with multiple tag or data arrays being simultaneously accessible
- G06F12/0848—Partitioned cache, e.g. separate instruction and operand caches
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F12/00—Accessing, addressing or allocating within memory systems or architectures
- G06F12/02—Addressing or allocation; Relocation
- G06F12/08—Addressing or allocation; Relocation in hierarchically structured memory systems, e.g. virtual memory systems
- G06F12/0802—Addressing of a memory level in which the access to the desired data or data block requires associative addressing means, e.g. caches
- G06F12/0806—Multiuser, multiprocessor or multiprocessing cache systems
- G06F12/0842—Multiuser, multiprocessor or multiprocessing cache systems for multiprocessing or multitasking
Definitions
- the invention relates to buffer memory management in a system having multiple execution entities.
- a buffer memory can be a relatively small, fast memory placed between a memory and another device that is capable of accessing the memory.
- An example of a buffer memory is a cache memory located between a processor and system memory (which typically is relatively large and slow) to reduce the effective access time required by the processor to retrieve information from the system memory.
- a multi-level cache system may be used for further performance improvement.
- a first-level cache (L1 cache) may be implemented in the processor itself, and a second-level, typically larger cache (L2 cache) is externally coupled to the processor.
- a cache memory may include separate instruction and data cache units, one to store instructions and the other to store data.
- a processor may fetch instructions from system memory to store in the instruction cache unit. Data processed by those instructions may be stored in the data cache unit. If information, such as instruction or data, requested by the processor is already stored in cache memory, then a cache memory hit is said to have occurred.
- a cache memory hit reduces the time needed for the processor to access information stored in memory, which improves processor performance.
- a cache miss is said to have occurred.
- the processor has to access the system memory to retrieve the desired information, which results in a memory access time performance reduction while the processor waits for the slower system memory to respond to the request.
- different cache management policies have been implemented.
- One of several mapping schemes may be selected, for example, including a direct mapping scheme or a set associative cache mapping scheme.
- One of several replacement policies may also be specified to improve cache memory hit ratios, including a first-in-first-out (FIFO) or least recently used (LRU) policy.
- FIFO first-in-first-out
- LRU least recently used
- Another feature of a cache memory that may be configured is the cache memory update policy that specifies how the system memory is updated when a write operation changes the contents of the cache. Update policies include a write-through policy or a write-back policy.
- a system such as a computer, may include multiple application programs and other software layers that have different data flow needs.
- a program execution entity such as a process, task, or thread, associated with a multimedia application may transfer large blocks of data (e.g., video data) that are typically not reused.
- access of these types of data may cause a cache to fill up with large blocks of data that are not likely to be reused.
- data used by one execution entity may replace data used by another execution entity, a phenomenon referred to as data cache pollution.
- Data cache pollution caused by the activities of one execution entity may increase the likelihood of cache misses for another execution entity, which may reduce overall system performance.
- a system includes a processor and a plurality of execution entities executable on the processor.
- a buffer memory in the system has multiple buffer sections. Each buffer section is adapted to store information associated with requests from a corresponding one of the multiple execution entities.
- FIG. 1 is a block diagram of portions of a buffer or cache memory having multiple sections according to an embodiment of the invention.
- FIG. 2 is a block diagram of an embodiment of a system including the cache memory of FIG. 1.
- FIG. 3 illustrates the components of each cache module in the cache memory of FIG. 1.
- FIG. 4 is a block diagram of a processor including the cache memory of FIG. 1 along with associated control logic.
- FIG. 5 is a flow diagram of an instruction execution sequence performed in the processor of FIG. 4.
- FIG. 6 is a flow diagram of an operating system in the system of FIG. 2 that sets up a cache memory according to an embodiment.
- Some embodiments of the invention include a system having a buffer memory that includes several individual buffer sections at one level of the memory hierarchy.
- Each buffer section may be a separate buffer module or may be a portion of a buffer memory that is separately addressable (that is, memory is separated into different address spaces).
- the individual buffer sections may be separately configurable and may be assigned to store information of different program execution entities in the system.
- Such a buffer memory may be referred to as a multi-unit buffer memory.
- the buffer memory may include a cache memory used in any of a variety of applications, e.g., processor subsystems, peripheral device controllers (such as video controllers, hard disk drive controllers, and so forth), and other types of control devices.
- Systems including such cache memories may include a general-purpose or special-purpose computer, a hand-held electronic device (e.g., telephones, calendar systems, electronic game devices, and the like), appliances, set-top boxes, and other electronics systems.
- a cache memory having multiple cache sections may be referred to as a multi-unit cache memory.
- a cache memory section may include a separate cache module or a portion of the cache memory that is separately addressable.
- the following described embodiments include a computer having a multi-unit cache memory with multiple independent cache modules —it is to be understood, however, that further embodiments may include computers having multi-unit cache memories with other independently configurable cache sections or other types of systems with buffer memories.
- each individual cache module in a multi-unit cache memory may be independently configurable.
- Such attributes may include each cache module's size, organization (e.g., direct mapped versus set associative mapping), replacement policy, update policy, and so forth.
- one cache module may be configured to be a direct mapped cache while another cache module may be configured as a k-way set associative cache.
- the cache modules may also be configured to have different update policies, including a write-through policy or a write-back policy.
- Other attributes may also be set differently for the different cache modules, as further described below.
- processors may be capable of receiving requests from multiple execution entities for processing.
- a processor may include, by way of example, a general-purpose or a special-purpose microprocessor, a microcontroller, or other types of control devices such as application-specific integrated circuits (ASICs), programmable gate arrays (PGAs), and the like.
- a program execution entity may be the basic unit of work of software and firmware layers that are loaded in the system. Such basic units of work may include processes, tasks, threads, or other units, as definable according to different systems. For example, in some operating systems, such as certain Windows® operating systems by Microsoft Corporation, multiple threads associated with processes in the system may be executable by the processor to perform different operations.
- Another operating system that offers multithreading or multitasking capabilities is the Be Operating System (BeOS) from BE, Inc., as described in the BE Operating System Product Data Sheet, published at http://www.be.com in 1998.
- BeOS Be Operating System
- multiple execution entities associated with different software and firmware layers may be active at a time. Requests from these execution entities are scheduled by the operating system according to a predetermined priority protocol, e.g., round-robin, etc.
- a predetermined priority protocol e.g., round-robin, etc.
- Such operating systems are said to be multitasking or multithreading operating systems.
- the independent cache modules of a multi-unit cache memory may be assigned to store information of corresponding execution entities.
- execution entities of a multimedia application may be assigned to one cache module, while execution entities of other applications may be assigned to different cache modules of the multi-unit cache memory.
- requests from each execution entity may be assigned to different execution entity identifiers (EIDs).
- EIDs execution entity identifiers
- a cache module may be configured for the general data usage behavior of an assigned application.
- the execution entities created by one software or firmware layer may further be subdivided to have multiple EIDs.
- an application may create execution entities that process data according to different temporal and spatial locality characteristics. For example, some execution entities may be more likely to reuse data than other execution entities created by the same application. Thus, it may be beneficial to further separately assign these different execution entities to different cache modules in the multi-unit cache memory.
- requests from different execution entities of one application may be assigned more than one EID so that different cache modules may be utilized.
- execution entities of different applications may be assigned the same EID.
- a first execution entity of a multimedia application may be assigned EID 1
- a second execution entity of the multimedia application may be assigned EID 2.
- execution entities of a spreadsheet application having similar data usage characteristics as the second execution entity of the multimedia application may also be assigned EID 2.
- a cache controller for the cache memory can keep track of which cache module of the multi-unit cache memory is to be used to store data accessed by the instruction.
- cache utilization may be improved since the individual cache modules may be configured to take advantage of the data usage characteristics of the different execution entities in the system.
- a multimedia application may typically generate requests that transfer large blocks of data that are not re-used.
- a cache module assigned to these types of requests may be configured to implement the FIFO replacement policy and write-through update policy.
- Cache modules assigned to other types of requests may have different configurations.
- EID identifiers may be assigned to these execution entities by an operating system.
- the operating system may access (at 504 ) configuration information loaded during system initialization to determine how EID identifiers are to be assigned.
- the operating system next assigns (at 506 ) the appropriate EID identifier to the execution entity.
- the operating system may be able to assign three EIDs to correspond to three cache modules in a multi-unit cache memory.
- Execution entities having one general data usage characteristic may be assigned a first EID identifier, and execution entities having a second general data usage characteristic may be assigned a second EID identifier.
- a default EID identifier may be assigned to those execution entities that are not specifically assigned one of the other two EID identifiers.
- the operating system also assigns (at 508 ) certain attributes of each cache module in the multi-unit cache memory. Such attributes may include the update, replacement, and placement policies. The operating system may also assign the attributes for the default cache module of the multi-unit cache memory. In alternative embodiments, the EID identifiers and cache attributes may be assigned as described above by a software layer that is separate from the operating system.
- execution entities of a multimedia application that transfer large amounts of data and that do not typically reuse the data may be allocated one EID identifier so that such data are stored in a first cache module configured for the cache data usage characteristics of these execution entities.
- Execution entities of intensive arithmetic applications such as compression applications, may be assigned another EID identifier so that data is stored in another cache module that is configured for cache data operations characterized by increased spatial locality.
- a multi-unit cache memory having multiple cache modules may be implemented in a multilevel cache memory having multiple levels of cache memory (e.g., an L1 cache and an L2 cache).
- Such a cache memory may be referred to as a multilevel, multi-unit cache memory, in which at least one level includes a multi-unit cache memory.
- a multilevel, multi-unit cache memory having two levels may be constructed in the following manner: the first level is a multi-unit cache and the second level is a conventional cache; the first level is a multi-unit cache and the second level is a multi-unit cache; or the first level is a conventional cache and the second level is a multi-unit cache.
- a multi-unit cache memory may include several P-caches, including a P0-cache, a P1-cache, a P2-cache, and so forth.
- the different P-caches may be implemented as separate memory elements or modules, e.g., multiple static random access memory (SRAM) or multiple dynamic random access memory (DRAM) devices.
- SRAM static random access memory
- DRAM dynamic random access memory
- multiple P-caches may be implemented in one memory device that is subdivided into separate sections to correspond to the different P-caches.
- the multi-unit cache memory may be integrated in another device, e.g., a processor or other control device in a system.
- the multi-unit cache memory may be a standalone unit accessible by control devices to retrieve cached data.
- one portion of the multi-unit cache memory may be located in one integrated device while another portion of the multi-unit cache memory is located in another device.
- each individual P-cache module in a multi-unit cache system may have different attributes, including cache size and organization and cache update, placement, and replacement policies.
- a placement policy may be specified for each P-cache to determine how information is placed into unfilled portions of the cache.
- a cache replacement policy is specified to manage replacement of information stored in each P-cache.
- Example replacement polices may include a first-in-first-out (FIFO) policy, a least-recently-used (LRU) policy, or some other type of replacement policy.
- a cache update policy manages how information is to be updated when a write occurs to the cache, which may include a write-through policy or a write-back policy.
- a multi-unit cache memory 100 includes several P-caches, shown as a P0-cache 102 , a P1-cache 104 , and a P2-cache 106 .
- a cache controller 108 is associated with the P0-, P1-, and P2-caches 102 , 104 , and 106 .
- separate address and data buses may be coupled to each of the cache modules 102 , 104 and 106 so that the cache modules may be accessed concurrently.
- a common address and data bus may be coupled to the cache modules.
- the cache controller 108 provides control signals to each of the P-cache modules 102 - 106 .
- the cache controller 108 includes storage elements 118 , in the form of registers or the like, that are programmable by the operating system to specify the EID identifiers associated with each of the P-caches.
- the cache controller 108 selects one of the P-caches based on a comparison of the EID provided by a request and the EID values stored in the storage elements 118 .
- the cache controller 108 also includes a replacement and update control block 120 to control the replacement and update policies of the three separate cache modules, as determined by control information programmed in the storage elements 118 .
- the storage elements 118 may be programmed to indicate a FIFO replacement policy for one P-cache and an LRU replacement policy for another P-cache.
- the cache controller 108 may also include a tag compare block 122 to compare the tag of an incoming request to the tag stored in the selected one or ones of the P-caches to determine if a cache hit has occurred. Further, if update of main memory 206 (FIG. 2) is needed, a write-back buffer 124 stores the cache line of one of the P-caches to transfer to main memory 206 or an L2 cache 204 (FIG. 2).
- the cache controller 108 also includes a cache coherency block 126 that determines if an accessed location of a cache module is valid.
- each cache module may store a valid/invalid bit.
- a more sophisticated coherency protocol may be implemented, such as the Modified, Exclusive, Shared, and Invalid (MESI) protocol.
- control signals that may be provided to the cache controller 108 may include a cache disable (CD) signal and a cache flush (CF) signal.
- CD cache disable
- CF cache flush
- other cache related signals such as snoop signals may be provided to the cache controller 108 .
- the multi-unit cache system 100 may be implemented in a number of different locations (e.g., processor subsystem, bridge controllers, peripheral device controllers, storage device controllers, and the like) in a system 10 .
- the system 10 includes a computer, although in alternative embodiments, the system 10 may be any other electronic device in which a cache or buffer memory may be implemented.
- the system 10 includes a central processing unit (CPU) 200 , which may include a processor or other suitable control device, having one or more levels of cache memory.
- the CPU 200 may include an internal cache that is the level-one (L1) cache 202 .
- the CPU 200 may be coupled over a host bus 203 to access an external cache that is the level-two (L2) cache 204 .
- the L1 cache 202 may include a code component (for storing instructions) and a data component (for storing data).
- the L2 cache 204 may include code and data components.
- instructions and data fetched from main memory 206 are stored in the code and data components, respectively, of the L1 or L2 cache 202 or 204 .
- separate code and data cache components are not implemented.
- the multi-unit cache memory 100 may be implemented in the L1 cache 202 , the L2 cache 204 , or both.
- the multi-unit cache memory 100 of FIG. 1 is implemented in the L1 cache 202 that is the internal cache of the CPU 200 . It is to be understood, however, that the multi-unit cache memory described, or modifications of such a cache memory, may be implemented in the L2 cache 204 or in other controllers in the system, such as a video controller or a hard disk drive controller, as examples.
- the multi-unit cache memory 100 forms the data cache component of the L1 cache 202 .
- the main memory 206 is controlled by a memory controller 207 in a memory hub 208 coupled to the CPU 200 over the host bus 203 .
- the memory hub 208 may include a cache controller 205 operatively coupled to the L2 cache 204 .
- the memory hub 208 may also include a graphics interface 211 that is coupled over a link 209 to a graphics controller 210 , which is in turn coupled to a display 212 .
- the graphics interface may be according to the Accelerated Graphics Port (A.G.P.) Interface Specification, Revision 2.0, published in May 1998.
- the memory hub 208 may also be coupled to an input/output (I/O) hub 214 that includes bridge controllers 215 and 223 coupled to a system bus 216 and a secondary bus 224 , respectively.
- the system bus may be a Peripheral Component Interconnect (PCI) bus, as defined by the PCI Local Bus Specification, Production Version, Revision 2.1, published in June 1995.
- the system bus 216 may be coupled to a storage controller 218 that controls access to one or more mass storage devices 220 , including a hard disk drive, a compact disc (CD) drive, or a digital video disc (DVD) drive.
- the storage controller 218 may be integrated into the I/O hub 214 , as may other control functions.
- the system bus 216 may also be coupled to other components, including, for example, a network controller 222 that is coupled to a network port (not shown).
- additional devices 226 may be coupled, as may be a non-volatile memory 228 that may store power up routines, such as basic input/output system (BIOS) routines.
- BIOS basic input/output system
- the secondary bus 224 may also include ports for coupling to peripheral devices.
- a host bridge controller and a system bridge controller may provide equivalent functions, with the host bridge controller coupled between the CPU 200 and the system bus 216 and the system bridge controller 224 coupled between the system bus 216 and the secondary bus 224 .
- any of a number of bus protocols may be implemented.
- Various different program execution entities are executable by the CPU 200 in the system 10 .
- multiple processes 252 , 254 , and 256 are loaded under an operating system 250 , which may be a Windows® operating system, for example.
- Each process may generate one or more execution entities that form the basic units of work in the system.
- the execution entities may be threads; as illustrated in FIG. 2, the process 252 may include threads 258 and 260 , the process 254 may include a thread 262 , and the process 256 may include threads 264 and 266 .
- Various software or firmware may be stored or otherwise tangibly embodied in one or more storage media in the system.
- Storage media suitable for tangibly embodying software and firmware instructions may include different forms of memory including semiconductor memory devices such as dynamic or static random access memories, erasable and programmable read-only memories (EPROMs), electrically erasable and programmable read-only memories (EEPROMs), and flash memories; magnetic disks such as fixed, floppy and removable disks; other magnetic media including tape; and optical media such as CD or DVD disks.
- the instructions stored in the storage media when executed cause the system 10 to perform programmed acts.
- the software or firmware can be loaded into the system 10 in one of many different ways. For example, instructions or other code segments stored on storage media or transported through a network interface card, modem, or other interface mechanism may be loaded into the system 10 and executed to perform programmed acts. In the loading or transport process, data signals that are embodied as carrier waves (transmitted over telephone lines, network lines, wireless links, cables and the like) may communicate the instructions or code segments to the system 10 .
- the execution entities are adapted to perform different operations. For example, a spreadsheet process may create a first thread to perform calculations on entries entered by a user and a second thread to transfer the calculated data into main memory 206 . Each thread or execution entity is able to generate requests, which are stored as instructions in main memory 206 . These instructions are fetched by the CPU 200 from main memory 206 for execution.
- an execution entity identifier may be assigned to each execution entity running in the system 10 .
- the EID of each execution entity may be assigned by the operating system.
- a scheduler 270 schedules requests from the execution entities for processing by the CPU 200
- the associated EID of each execution entity is stored along with one or more corresponding instructions.
- the CPU 200 fetches the associated EIDs along with the instructions.
- the EIDs are not stored into memory 206 along with instructions. Instead, multiple instruction memory regions may be defined in the memory 206 to correspond to the different EIDs. Instructions associated with a request from an execution entity having a first EID may be stored in a first instruction memory region; instructions associated with a request from an execution entity having a second EID may be stored in a second instruction memory region; and so forth.
- the CPU 200 fetches instructions from memory 206 without associated EIDs. However, based on which of the instruction memory regions the instruction is fetched, the CPU 200 can determine the EID of the instruction.
- the CPU 200 may include multiple microsequencers assigned to different threads.
- one microsequencer may retrieve instructions associated with one thread, another microsequencer may retrieve instructions associated with another thread, and so forth.
- Each microsequencer may be configured to know locations of instructions of corresponding execution entities.
- an instruction's EID may be determined depending on which of the microsequencers fetched that instruction. The determined instruction may then be stored inside the CPU.
- the retrieved or determined EID is decoded by the cache controller 108 or by some other suitable decoder to identify which P-cache is to be used when the instruction requests an access to data.
- the cache controller 108 accesses one of the P-caches to retrieve or store data processed by the corresponding instruction.
- data associated with instructions having EID 0 may be stored in the P0-cache 102
- data associated with instructions having EID 1 may be stored in the P1-cache 104
- data associated with instructions having EID 2 may be stored in the P2-cache 106 .
- a P-cache may be associated with more than one EID.
- execution entities from different application and software layers may be assigned the same EID.
- Each P-cache may include a status array 160 , a tag array 162 , and a data array 164 . As illustrated, each of the status array 160 , tag array 162 , and data array 164 is divided into 4 different sections for the 4-way set-associative organization.
- the status array 160 may contain one or more of the following fields: an EID identifier; replacement selection bits (RPS) that are used by the replacement and update control block 120 to replace a cache line; and cache coherency protocol bits.
- each block of the P-cache module may be associated with a valid/invalid bit to indicate if the corresponding cache location is valid or invalid.
- the status array 160 may store MESI bits.
- the replacement selection bits RPS may be used to indicate which cache line is to be replaced.
- the RPS bits may be used to keep track of the least recently used cache line (for LRU replacement) or the first entered line (for FIFO replacement), as examples.
- the cache controller 108 may be implemented as an integrated unit or as several separate control units. As discussed, when an instruction is fetched for execution, the EID associated with the instruction is retrieved. Based on the EID value, the appropriate one of the P-cache modules is selected to retrieve data from or write data to. A hit or miss may be returned depending on whether a valid copy of the associated data is stored in the selected P-cache module.
- a multi-unit cache system having independently configurable cache modules may have one or more of the following advantages. Greater cache management flexibility may be available, since the placement, replacement, and update policies and cache size and organization of each of the P-cache modules may be set to improve cache utilization for corresponding execution entities. Cache performance may be improved by configuring cache modules to take advantage of different cache usage characteristics (to store data or instructions) of different execution entities. Data cache pollution by the different active execution entities in the system 10 may be reduced, which may improve the cache hit ratio.
- the multi-unit data cache system may offer high access bandwidth by increasing parallelism for a multithreading or multitasking processor since the P-cache modules may be concurrently accessible. Such concurrent data cache accesses may help reduce data cache latency to help meet the data access bandwidth demands of high-performance processors.
- compilers for different application programs may dynamically reconfigure attributes of the multi-unit cache memory to further enhance cache performance. For example, during operation, statistical information associated with different execution entities may be collected and stored. Depending on the collected statistical information, the attributes of each P-cache module may be changed. Thus, for example, if a FIFO replacement policy is determined not to be efficient for a particular P-cache module, the cache controller 108 may be notified to change the replacement policy to the LRU policy or some other replacement policy.
- This alternative embodiment may provide the flexibility of dynamically changing the configuration of individual P-cache modules in response to how execution entities in the system 10 are performing.
- the CPU 200 includes the multi-unit L1 cache memory 202 and associated logic.
- the multi-unit L1 cache memory includes the three data cache modules: the P0-, P1-, and P2-caches 102 , 104 , and 106 .
- the P0-cache 102 may be designated as the default data cache that is used to store data associated with execution entities that have not specifically been assigned to one of the other P-caches in the L1 cache. For example, such execution entities may be assigned a default EID 0 by the operating system.
- the P1- and P2-caches 104 and 106 may be assigned to store data for requests from execution entities having EIDs 1 and 2, respectively.
- the P0-cache may be a larger memory than either the P1- or P2-cache since it is the default data cache.
- a bus front unit (BFU) 404 forms the interface to the front side or host bus 203 .
- the BFU 404 may include address drivers and receivers, write buffers, data bus transceivers, bus control logic, bus master control, and parity generation and control.
- Instructions retrieved by the BFU 404 from either the main memory 206 or from the L2 cache 204 may be stored in an instruction cache 406 that is part of the L1 cache 202 .
- the internal instruction cache 406 may keep copies of the most frequently used instructions.
- instructions are fetched along with EIDs from either the main memory 206 or the L2 cache 204 and stored in the instruction cache 406 .
- An instruction buffer and decode logic 408 decodes a selected instruction and associated EID from the instruction cache 406 and produces one or more micro-operations along with corresponding EIDs.
- the instructions are stored in different instruction memory regions of the memory 206 according to different EIDs.
- the EIDs are not stored along with the instructions.
- the CPU 200 fetches an instruction, an associated EID is not retrieved. Instead, the CPU 200 determines the EID of the fetched instruction based on the address location where the instruction is stored. This may be performed, for example, by the decode logic 408 .
- the EID of an instruction is determined based on which instruction memory region the instruction is fetched from. Once the EID is determined by the CPU 200 , it can be attached to the decoded micro-operations and stored in the instruction queue 412 .
- multiple program counters and microsequencers may be included in the CPU 200 that are assigned to corresponding threads. This embodiment is described further below.
- the output port of the instruction buffer and decode logic 408 may be coupled to an instruction queue 412 , which stores the micro-operations along with associated EIDs.
- the output port of the instruction queue 412 is routed to a sequencer 414 .
- the sequencer 414 may include multiple microsequencer units 430 , 432 , and 434 corresponding to the different EIDs.
- the microsequencer unit 430 may be configured to process micro-operations associated with EID 0
- the microsequencer 432 may be configured to process micro-operations associated with EID 1
- the microsequencer 434 may be configured to process micro-operations associated with EID 2.
- micro-operations processed by the microsequencers 430 , 432 , and 434 are received from the instruction queue 412 .
- the microsequencers 430 , 432 , and 434 may operate simultaneously to process micro-operations associated with different EIDs. Operation of the microsequencers 430 , 432 , and 434 are controlled by a control logic 436 in the sequencer 414 .
- the EIDs associated with micro-operations are originally retrieved from memory along with instructions of the different execution entities.
- each microsequencer may be independently configured to fetch instructions of corresponding execution entities.
- a first microsequencer fetches instructions associated with a first execution entity
- a second microsequencer fetches instructions associated with a second execution entity, and so forth.
- the EIDs of a fetched instruction may thus be determined based on which of the microsequencers fetched the instruction.
- An execution entity typically includes a number of instructions that are executed in some program order. By default, instruction addresses are simply incremented to fetch the next instruction. If a jump or other conditional branch occurs, then a target address is specified for the address of the next instruction. Thus, the address of the memory location where the next instruction is stored is known.
- a program counter may be used to keep track of the program order of instructions.
- a microsequencer works in conjunction with the program counter to execute instructions. To fetch an instruction, the microsequencer may ask a fetch unit, located in the BFU 404 for example, to fetch an instruction with an address stored in the program counter. Thus, fetched instructions may be identified as belonging to an execution entity since the microsequencer already knows (from the program counter) the address of the next instruction.
- two or more independent program counters may be used.
- three program counters PC0, PC1, and PC2 may be associated with microsequencers 430 , 432 , and 434 , respectively.
- the operating system may load the initial states of the program counters PC0, PC1, and PC2 so that the program counters may fetch instructions associated with the different threads.
- the combination of PC0 and microsequencer 430 keeps track of the program sequence for a first thread
- the combination of PC1 and microsequencer 432 keeps track of the program sequence for a second thread, and so forth.
- the CPU 200 When an instruction pointed to by PC0 is fetched, the CPU 200 knows that the instruction belongs to a first thread having, for example, EID 0. The EID is then attached to the instruction in the CPU 200 and to subsequently decoded micro-operations that are stored in the instruction queue 412 for execution by the microsequencer 430 , 432 , or 434 in the sequencer 414 .
- the output port of the sequencer 414 is provided to a pipeline back-end block 415 that includes various functional units, such as for example, an early branch execution unit 416 , a fast decoder unit 418 , an arithmetic/logic unit (ALU) 420 , and an address generator unit (AGU) 422 .
- these functional units may be accessed to perform requested functions.
- the pipeline back-end block 415 also includes register files 424 , 426 , and 428 .
- the register files 424 , 426 and 428 in the CPU 200 correspond to the three EID groups EID 0, EID 1, and EID 2.
- the register files 424 , 426 and 428 may each include control registers, status registers, flag registers, and general purpose registers.
- the register files 424 , 426 , and 428 are updated by the functional units in the pipeline back-end block 415 during operation. According to one embodiment, the register files 424 , 426 , and 428 may also be accessible independently and concurrently.
- requests associated with different EIDs may be processed concurrently provided that there are no dependencies among the requests, and further, the multiple requests do not need to utilize the same functional units 416 , 418 , 420 , and 422 .
- the register files 424 , 426 , and 428 as well as the cache modules in the multi-unit cache memory may be accessed and updated concurrently.
- a store buffer 450 (for write operations) and a load buffer 452 (for read operations) store data that are retrieved from or targeted for the BFU 404 .
- the store and load buffers 450 and 452 are coupled to an internal data bus 454 that is coupled to several units, including the P0-cache 102 , the P1-cache 104 , the P2-cache 106 , the pipeline back-end block 415 , and a translation look aside buffer (TLB) 456 .
- TLB translation look aside buffer
- TLB 456 which is basically a high speed memory in the CPU 200 that translates the virtual address from the instruction cache 406 into a physical address to access the data cache modules 102 , 104 and 106 .
- control logic 436 in the microcode sequencer 414 may select an appropriate instruction for processing by one of the microsequencers 430 , 432 , and 434 . If data access is needed, the microsequencers 430 , 432 , and 434 may concurrently access the several modules in the multi-unit data cache. Thus, to improve system performance, multiple instructions may be executed in the CPU 200 with concurrent access to data in the L1 multi-unit cache 202 .
- control logic 436 of the sequencer 414 may also consider possible load/store ordering, outstanding data cache refilling, and other issues. For example, in one embodiment, instructions associated with a request that has been determined to have a high hit ratio may be scheduled first, as may instructions of a real-time constrained execution entity having high priority.
- Instructions are fetched from main memory 206 or L2 cache 204 (at 302 ) by the CPU 200 over the host bus 203 .
- associated EIDs are retrieved with the instructions.
- the associated EIDs are not stored and thus are not retrieved.
- the fetched instructions are then translated into internal micro-operations (at 304 ) by the decoder stage 408 , with a corresponding EID attached to each micro-operation.
- the EID may be the one fetched with the instruction or it may be determined by the CPU 200 based on the address location of the instruction or which microsequencer fetched the instruction.
- the translated micro-operation is stored in the instruction queue 412 (at 306 ).
- the micro-operation is then delivered to one of the microsequencers 430 , 432 , and 434 for execution (at 308 ).
- Execution of the micro-operation may cause a data cache access request to be made (at 310 ), in which case a corresponding one of the P-cache modules is accessed based on the attached EID.
- the EID is decoded by the cache controller 108 and an appropriate request is sent to a corresponding P-cache ( 102 , 104 , or 106 ).
- the data access request is then completed in the assigned P-cache (at 312 ).
- the embodiments described include a multi-unit cache memory to store data, it is contemplated that the multi-unit cache memory may be adapted to store instructions of different execution entities in further embodiments. In such embodiments, the information stored in the multi-unit cache memory includes the instructions themselves.
Abstract
A system includes multiple program execution entities (e.g., tasks, processes, threads, and the like) and a cache memory having multiple sections. An identifier is assigned to each execution entity. An instruction of one of the execution entities is retrieved and an associated identifier is decoded. Information associated with the instruction is stored in one of the cache sections based on the identifier.
Description
- The invention relates to buffer memory management in a system having multiple execution entities.
- A buffer memory can be a relatively small, fast memory placed between a memory and another device that is capable of accessing the memory. An example of a buffer memory is a cache memory located between a processor and system memory (which typically is relatively large and slow) to reduce the effective access time required by the processor to retrieve information from the system memory. In some systems, a multi-level cache system may be used for further performance improvement. A first-level cache (L1 cache) may be implemented in the processor itself, and a second-level, typically larger cache (L2 cache) is externally coupled to the processor.
- Further, in some conventional memory systems, a cache memory may include separate instruction and data cache units, one to store instructions and the other to store data. During operation, a processor may fetch instructions from system memory to store in the instruction cache unit. Data processed by those instructions may be stored in the data cache unit. If information, such as instruction or data, requested by the processor is already stored in cache memory, then a cache memory hit is said to have occurred. A cache memory hit reduces the time needed for the processor to access information stored in memory, which improves processor performance.
- However, if information needed by the processor is not stored in cache memory, then a cache miss is said to have occurred. When a cache miss occurs, the processor has to access the system memory to retrieve the desired information, which results in a memory access time performance reduction while the processor waits for the slower system memory to respond to the request. To reduce cache misses, different cache management policies have been implemented. One of several mapping schemes may be selected, for example, including a direct mapping scheme or a set associative cache mapping scheme. A set associative cache memory that implements k-way associative mapping, e.g., 2-way associative mapping, 4-way associative mapping, and so forth, generally provides a higher hit ratio than direct mapped cache memory. One of several replacement policies may also be specified to improve cache memory hit ratios, including a first-in-first-out (FIFO) or least recently used (LRU) policy. Another feature of a cache memory that may be configured is the cache memory update policy that specifies how the system memory is updated when a write operation changes the contents of the cache. Update policies include a write-through policy or a write-back policy.
- Conventionally, a system, such as a computer, may include multiple application programs and other software layers that have different data flow needs. For example, a program execution entity, such as a process, task, or thread, associated with a multimedia application may transfer large blocks of data (e.g., video data) that are typically not reused. Thus, access of these types of data may cause a cache to fill up with large blocks of data that are not likely to be reused.
- In filling a cache memory, data used by one execution entity may replace data used by another execution entity, a phenomenon referred to as data cache pollution. Data cache pollution caused by the activities of one execution entity may increase the likelihood of cache misses for another execution entity, which may reduce overall system performance.
- A need thus exists for a memory architecture that provides improved memory performance.
- In general, according to an embodiment, a system includes a processor and a plurality of execution entities executable on the processor. A buffer memory in the system has multiple buffer sections. Each buffer section is adapted to store information associated with requests from a corresponding one of the multiple execution entities.
- Other features will become apparent from the following description and from the claims.
- FIG. 1 is a block diagram of portions of a buffer or cache memory having multiple sections according to an embodiment of the invention.
- FIG. 2 is a block diagram of an embodiment of a system including the cache memory of FIG. 1.
- FIG. 3 illustrates the components of each cache module in the cache memory of FIG. 1.
- FIG. 4 is a block diagram of a processor including the cache memory of FIG. 1 along with associated control logic.
- FIG. 5 is a flow diagram of an instruction execution sequence performed in the processor of FIG. 4.
- FIG. 6 is a flow diagram of an operating system in the system of FIG. 2 that sets up a cache memory according to an embodiment.
- In the following description, numerous details are set forth to provide an understanding of the present invention. However, it is to be understood by those skilled in the art that the present invention may be practiced without these details and that numerous variations or modifications from the described embodiments may be possible.
- Some embodiments of the invention include a system having a buffer memory that includes several individual buffer sections at one level of the memory hierarchy. Each buffer section may be a separate buffer module or may be a portion of a buffer memory that is separately addressable (that is, memory is separated into different address spaces). The individual buffer sections may be separately configurable and may be assigned to store information of different program execution entities in the system. Such a buffer memory may be referred to as a multi-unit buffer memory.
- In some embodiments, the buffer memory may include a cache memory used in any of a variety of applications, e.g., processor subsystems, peripheral device controllers (such as video controllers, hard disk drive controllers, and so forth), and other types of control devices. Systems including such cache memories may include a general-purpose or special-purpose computer, a hand-held electronic device (e.g., telephones, calendar systems, electronic game devices, and the like), appliances, set-top boxes, and other electronics systems. A cache memory having multiple cache sections may be referred to as a multi-unit cache memory. A cache memory section may include a separate cache module or a portion of the cache memory that is separately addressable. The following described embodiments include a computer having a multi-unit cache memory with multiple independent cache modules —it is to be understood, however, that further embodiments may include computers having multi-unit cache memories with other independently configurable cache sections or other types of systems with buffer memories.
- According to some embodiments, the attributes of each individual cache module in a multi-unit cache memory may be independently configurable. Such attributes may include each cache module's size, organization (e.g., direct mapped versus set associative mapping), replacement policy, update policy, and so forth. Thus, for example, one cache module may be configured to be a direct mapped cache while another cache module may be configured as a k-way set associative cache. The cache modules may also be configured to have different update policies, including a write-through policy or a write-back policy. Other attributes may also be set differently for the different cache modules, as further described below.
- Some processors may be capable of receiving requests from multiple execution entities for processing. A processor may include, by way of example, a general-purpose or a special-purpose microprocessor, a microcontroller, or other types of control devices such as application-specific integrated circuits (ASICs), programmable gate arrays (PGAs), and the like. A program execution entity according to one embodiment may be the basic unit of work of software and firmware layers that are loaded in the system. Such basic units of work may include processes, tasks, threads, or other units, as definable according to different systems. For example, in some operating systems, such as certain Windows® operating systems by Microsoft Corporation, multiple threads associated with processes in the system may be executable by the processor to perform different operations. Another operating system that offers multithreading or multitasking capabilities is the Be Operating System (BeOS) from BE, Inc., as described in the BE Operating System Product Data Sheet, published at http://www.be.com in 1998.
- In such operating systems, multiple execution entities associated with different software and firmware layers may be active at a time. Requests from these execution entities are scheduled by the operating system according to a predetermined priority protocol, e.g., round-robin, etc. Such operating systems are said to be multitasking or multithreading operating systems. To take advantage of the multitasking or multithreading capabilities of a system, the independent cache modules of a multi-unit cache memory may be assigned to store information of corresponding execution entities. Thus, for example, execution entities of a multimedia application may be assigned to one cache module, while execution entities of other applications may be assigned to different cache modules of the multi-unit cache memory. To that end, according to one embodiment, requests from each execution entity may be assigned to different execution entity identifiers (EIDs). Thus, requests from execution entities of a first application may be assigned to one EID, and requests from another execution entity may be assigned another EID. Thus, according to this embodiment, a cache module may be configured for the general data usage behavior of an assigned application.
- In another embodiment, the execution entities created by one software or firmware layer may further be subdivided to have multiple EIDs. For example, an application may create execution entities that process data according to different temporal and spatial locality characteristics. For example, some execution entities may be more likely to reuse data than other execution entities created by the same application. Thus, it may be beneficial to further separately assign these different execution entities to different cache modules in the multi-unit cache memory. Thus, in one alternative embodiment, requests from different execution entities of one application may be assigned more than one EID so that different cache modules may be utilized. In addition, execution entities of different applications may be assigned the same EID. Thus, for example, a first execution entity of a multimedia application may be assigned
EID 1, while a second execution entity of the multimedia application may be assigned EID 2. In the same system, execution entities of a spreadsheet application having similar data usage characteristics as the second execution entity of the multimedia application may also be assigned EID 2. - In further embodiments, other different schemes may be implemented in assigning EIDs to requests of execution entities. Based on the EID associated with an instruction, a cache controller for the cache memory can keep track of which cache module of the multi-unit cache memory is to be used to store data accessed by the instruction. As a result, cache utilization may be improved since the individual cache modules may be configured to take advantage of the data usage characteristics of the different execution entities in the system. For example, a multimedia application may typically generate requests that transfer large blocks of data that are not re-used. A cache module assigned to these types of requests may be configured to implement the FIFO replacement policy and write-through update policy. Cache modules assigned to other types of requests may have different configurations.
- As execution entities are created in a system, EID identifiers may be assigned to these execution entities by an operating system. Referring to FIG. 6, according to one embodiment, if a new execution entity is detected (at502), the operating system may access (at 504) configuration information loaded during system initialization to determine how EID identifiers are to be assigned. The operating system next assigns (at 506) the appropriate EID identifier to the execution entity. For example, the operating system may be able to assign three EIDs to correspond to three cache modules in a multi-unit cache memory. Execution entities having one general data usage characteristic may be assigned a first EID identifier, and execution entities having a second general data usage characteristic may be assigned a second EID identifier. A default EID identifier may be assigned to those execution entities that are not specifically assigned one of the other two EID identifiers.
- In addition, based on the configuration information, the operating system also assigns (at508) certain attributes of each cache module in the multi-unit cache memory. Such attributes may include the update, replacement, and placement policies. The operating system may also assign the attributes for the default cache module of the multi-unit cache memory. In alternative embodiments, the EID identifiers and cache attributes may be assigned as described above by a software layer that is separate from the operating system.
- In one example, execution entities of a multimedia application that transfer large amounts of data and that do not typically reuse the data may be allocated one EID identifier so that such data are stored in a first cache module configured for the cache data usage characteristics of these execution entities. Execution entities of intensive arithmetic applications, such as compression applications, may be assigned another EID identifier so that data is stored in another cache module that is configured for cache data operations characterized by increased spatial locality.
- In some embodiments, a multi-unit cache memory having multiple cache modules may be implemented in a multilevel cache memory having multiple levels of cache memory (e.g., an L1 cache and an L2 cache). Such a cache memory may be referred to as a multilevel, multi-unit cache memory, in which at least one level includes a multi-unit cache memory. Thus, for example, a multilevel, multi-unit cache memory having two levels may be constructed in the following manner: the first level is a multi-unit cache and the second level is a conventional cache; the first level is a multi-unit cache and the second level is a multi-unit cache; or the first level is a conventional cache and the second level is a multi-unit cache.
- The individual cache modules of a multi-unit cache may be referred to as P-caches. Thus, for example, a multi-unit cache memory may include several P-caches, including a P0-cache, a P1-cache, a P2-cache, and so forth. The different P-caches may be implemented as separate memory elements or modules, e.g., multiple static random access memory (SRAM) or multiple dynamic random access memory (DRAM) devices. Alternatively, multiple P-caches may be implemented in one memory device that is subdivided into separate sections to correspond to the different P-caches. In addition, the multi-unit cache memory may be integrated in another device, e.g., a processor or other control device in a system. Alternatively, the multi-unit cache memory may be a standalone unit accessible by control devices to retrieve cached data. In further embodiments, one portion of the multi-unit cache memory may be located in one integrated device while another portion of the multi-unit cache memory is located in another device.
- In some embodiments of the invention, each individual P-cache module in a multi-unit cache system may have different attributes, including cache size and organization and cache update, placement, and replacement policies. A placement policy may be specified for each P-cache to determine how information is placed into unfilled portions of the cache. A cache replacement policy is specified to manage replacement of information stored in each P-cache. Example replacement polices may include a first-in-first-out (FIFO) policy, a least-recently-used (LRU) policy, or some other type of replacement policy. A cache update policy manages how information is to be updated when a write occurs to the cache, which may include a write-through policy or a write-back policy.
- Referring to FIG. 1, a
multi-unit cache memory 100 according to an embodiment includes several P-caches, shown as a P0-cache 102, a P1-cache 104, and a P2-cache 106. Acache controller 108 is associated with the P0-, P1-, and P2-caches cache modules cache controller 108 provides control signals to each of the P-cache modules 102-106. - The
cache controller 108 includesstorage elements 118, in the form of registers or the like, that are programmable by the operating system to specify the EID identifiers associated with each of the P-caches. When themulti-unit cache memory 100 is accessed, thecache controller 108 selects one of the P-caches based on a comparison of the EID provided by a request and the EID values stored in thestorage elements 118. - The
cache controller 108 also includes a replacement and update control block 120 to control the replacement and update policies of the three separate cache modules, as determined by control information programmed in thestorage elements 118. Thus, for example, thestorage elements 118 may be programmed to indicate a FIFO replacement policy for one P-cache and an LRU replacement policy for another P-cache. - The
cache controller 108 may also include a tag compareblock 122 to compare the tag of an incoming request to the tag stored in the selected one or ones of the P-caches to determine if a cache hit has occurred. Further, if update of main memory 206 (FIG. 2) is needed, a write-back buffer 124 stores the cache line of one of the P-caches to transfer tomain memory 206 or an L2 cache 204 (FIG. 2). - To ensure cache data integrity, the
cache controller 108 also includes acache coherency block 126 that determines if an accessed location of a cache module is valid. In one embodiment, each cache module may store a valid/invalid bit. Alternatively, a more sophisticated coherency protocol may be implemented, such as the Modified, Exclusive, Shared, and Invalid (MESI) protocol. - Other control signals that may be provided to the
cache controller 108 may include a cache disable (CD) signal and a cache flush (CF) signal. In addition, other cache related signals such as snoop signals may be provided to thecache controller 108. - Referring to FIG. 2, the
multi-unit cache system 100 may be implemented in a number of different locations (e.g., processor subsystem, bridge controllers, peripheral device controllers, storage device controllers, and the like) in asystem 10. In one embodiment, thesystem 10 includes a computer, although in alternative embodiments, thesystem 10 may be any other electronic device in which a cache or buffer memory may be implemented. - The
system 10 includes a central processing unit (CPU) 200, which may include a processor or other suitable control device, having one or more levels of cache memory. For example, as illustrated, theCPU 200 may include an internal cache that is the level-one (L1)cache 202. In addition, theCPU 200 may be coupled over ahost bus 203 to access an external cache that is the level-two (L2)cache 204. TheL1 cache 202 may include a code component (for storing instructions) and a data component (for storing data). Similarly, theL2 cache 204 may include code and data components. Thus, instructions and data fetched frommain memory 206 are stored in the code and data components, respectively, of the L1 orL2 cache - In some embodiments, the multi-unit cache memory100 (FIG. 1) may be implemented in the
L1 cache 202, theL2 cache 204, or both. For purposes of this discussion, it is assumed that themulti-unit cache memory 100 of FIG. 1 is implemented in theL1 cache 202 that is the internal cache of theCPU 200. It is to be understood, however, that the multi-unit cache memory described, or modifications of such a cache memory, may be implemented in theL2 cache 204 or in other controllers in the system, such as a video controller or a hard disk drive controller, as examples. In addition, in this embodiment, themulti-unit cache memory 100 forms the data cache component of theL1 cache 202. - The
main memory 206 is controlled by amemory controller 207 in amemory hub 208 coupled to theCPU 200 over thehost bus 203. In addition, thememory hub 208 may include acache controller 205 operatively coupled to theL2 cache 204. Thememory hub 208 may also include agraphics interface 211 that is coupled over alink 209 to agraphics controller 210, which is in turn coupled to adisplay 212. As an example, the graphics interface may be according to the Accelerated Graphics Port (A.G.P.) Interface Specification, Revision 2.0, published in May 1998. - The
memory hub 208 may also be coupled to an input/output (I/O)hub 214 that includesbridge controllers system bus 216 and asecondary bus 224, respectively. As an example, the system bus may be a Peripheral Component Interconnect (PCI) bus, as defined by the PCI Local Bus Specification, Production Version, Revision 2.1, published in June 1995. Thesystem bus 216 may be coupled to astorage controller 218 that controls access to one or moremass storage devices 220, including a hard disk drive, a compact disc (CD) drive, or a digital video disc (DVD) drive. In an alternative embodiment, thestorage controller 218 may be integrated into the I/O hub 214, as may other control functions. Thesystem bus 216 may also be coupled to other components, including, for example, anetwork controller 222 that is coupled to a network port (not shown). - On the
secondary bus 224,additional devices 226 may be coupled, as may be anon-volatile memory 228 that may store power up routines, such as basic input/output system (BIOS) routines. Thesecondary bus 224 may also include ports for coupling to peripheral devices. Although the description makes reference to specific configurations and architectures of the various layers of thesystem 10, it is contemplated that numerous modifications and variations of the described and illustrated embodiments may be possible. For example, instead of memory and I/O hubs, a host bridge controller and a system bridge controller may provide equivalent functions, with the host bridge controller coupled between theCPU 200 and thesystem bus 216 and thesystem bridge controller 224 coupled between thesystem bus 216 and thesecondary bus 224. In addition, any of a number of bus protocols may be implemented. - Various different program execution entities are executable by the
CPU 200 in thesystem 10. As illustrated, according to one embodiment,multiple processes operating system 250, which may be a Windows® operating system, for example. Each process may generate one or more execution entities that form the basic units of work in the system. In one example, the execution entities may be threads; as illustrated in FIG. 2, theprocess 252 may includethreads process 254 may include athread 262, and theprocess 256 may includethreads - Various software or firmware (formed of modules, routines, or other layers, for example), including applications, operating system modules or routines, device drivers, BIOS modules or routines, and interrupt handlers, may be stored or otherwise tangibly embodied in one or more storage media in the system. Storage media suitable for tangibly embodying software and firmware instructions may include different forms of memory including semiconductor memory devices such as dynamic or static random access memories, erasable and programmable read-only memories (EPROMs), electrically erasable and programmable read-only memories (EEPROMs), and flash memories; magnetic disks such as fixed, floppy and removable disks; other magnetic media including tape; and optical media such as CD or DVD disks. The instructions stored in the storage media when executed cause the
system 10 to perform programmed acts. - The software or firmware can be loaded into the
system 10 in one of many different ways. For example, instructions or other code segments stored on storage media or transported through a network interface card, modem, or other interface mechanism may be loaded into thesystem 10 and executed to perform programmed acts. In the loading or transport process, data signals that are embodied as carrier waves (transmitted over telephone lines, network lines, wireless links, cables and the like) may communicate the instructions or code segments to thesystem 10. - The execution entities (in this case threads) are adapted to perform different operations. For example, a spreadsheet process may create a first thread to perform calculations on entries entered by a user and a second thread to transfer the calculated data into
main memory 206. Each thread or execution entity is able to generate requests, which are stored as instructions inmain memory 206. These instructions are fetched by theCPU 200 frommain memory 206 for execution. - According to some embodiments, an execution entity identifier (EID) may be assigned to each execution entity running in the
system 10. The EID of each execution entity may be assigned by the operating system. In one embodiment, when ascheduler 270 schedules requests from the execution entities for processing by theCPU 200, the associated EID of each execution entity is stored along with one or more corresponding instructions. In this embodiment, theCPU 200 fetches the associated EIDs along with the instructions. - In an alternative embodiment, the EIDs are not stored into
memory 206 along with instructions. Instead, multiple instruction memory regions may be defined in thememory 206 to correspond to the different EIDs. Instructions associated with a request from an execution entity having a first EID may be stored in a first instruction memory region; instructions associated with a request from an execution entity having a second EID may be stored in a second instruction memory region; and so forth. In this alternative embodiment, theCPU 200 fetches instructions frommemory 206 without associated EIDs. However, based on which of the instruction memory regions the instruction is fetched, theCPU 200 can determine the EID of the instruction. - In yet a further embodiment, in which EIDs are similarly not stored along with instructions, the
CPU 200 may include multiple microsequencers assigned to different threads. Thus, one microsequencer may retrieve instructions associated with one thread, another microsequencer may retrieve instructions associated with another thread, and so forth. Each microsequencer may be configured to know locations of instructions of corresponding execution entities. In this embodiment, an instruction's EID may be determined depending on which of the microsequencers fetched that instruction. The determined instruction may then be stored inside the CPU. - The retrieved or determined EID is decoded by the
cache controller 108 or by some other suitable decoder to identify which P-cache is to be used when the instruction requests an access to data. Thecache controller 108 accesses one of the P-caches to retrieve or store data processed by the corresponding instruction. With the example configuration of FIG. 1, data associated with instructions having EID 0 may be stored in the P0-cache 102, data associated withinstructions having EID 1 may be stored in the P1-cache 104, and data associated with instructions having EID 2 may be stored in the P2-cache 106. In some embodiments, a P-cache may be associated with more than one EID. Further, execution entities from different application and software layers may be assigned the same EID. - Referring to FIG. 3, the general architecture of one of the P-caches is illustrated. In the example shown in FIG. 3, a 4-way set-associative cache is illustrated. Other configurations are also possible, including a direct mapped cache or other k-way set-associative caches. Each P-cache may include a
status array 160, atag array 162, and adata array 164. As illustrated, each of thestatus array 160,tag array 162, anddata array 164 is divided into 4 different sections for the 4-way set-associative organization. - The
status array 160 may contain one or more of the following fields: an EID identifier; replacement selection bits (RPS) that are used by the replacement and update control block 120 to replace a cache line; and cache coherency protocol bits. For example, each block of the P-cache module may be associated with a valid/invalid bit to indicate if the corresponding cache location is valid or invalid. Alternatively, thestatus array 160 may store MESI bits. The replacement selection bits RPS may be used to indicate which cache line is to be replaced. The RPS bits may be used to keep track of the least recently used cache line (for LRU replacement) or the first entered line (for FIFO replacement), as examples. - The
cache controller 108 may be implemented as an integrated unit or as several separate control units. As discussed, when an instruction is fetched for execution, the EID associated with the instruction is retrieved. Based on the EID value, the appropriate one of the P-cache modules is selected to retrieve data from or write data to. A hit or miss may be returned depending on whether a valid copy of the associated data is stored in the selected P-cache module. - A multi-unit cache system having independently configurable cache modules according to some embodiments may have one or more of the following advantages. Greater cache management flexibility may be available, since the placement, replacement, and update policies and cache size and organization of each of the P-cache modules may be set to improve cache utilization for corresponding execution entities. Cache performance may be improved by configuring cache modules to take advantage of different cache usage characteristics (to store data or instructions) of different execution entities. Data cache pollution by the different active execution entities in the
system 10 may be reduced, which may improve the cache hit ratio. In addition, the multi-unit data cache system may offer high access bandwidth by increasing parallelism for a multithreading or multitasking processor since the P-cache modules may be concurrently accessible. Such concurrent data cache accesses may help reduce data cache latency to help meet the data access bandwidth demands of high-performance processors. - In another embodiment, compilers for different application programs may dynamically reconfigure attributes of the multi-unit cache memory to further enhance cache performance. For example, during operation, statistical information associated with different execution entities may be collected and stored. Depending on the collected statistical information, the attributes of each P-cache module may be changed. Thus, for example, if a FIFO replacement policy is determined not to be efficient for a particular P-cache module, the
cache controller 108 may be notified to change the replacement policy to the LRU policy or some other replacement policy. This alternative embodiment may provide the flexibility of dynamically changing the configuration of individual P-cache modules in response to how execution entities in thesystem 10 are performing. - Referring to FIG. 4, in one embodiment, the
CPU 200 includes the multi-unitL1 cache memory 202 and associated logic. The multi-unit L1 cache memory includes the three data cache modules: the P0-, P1-, and P2-caches cache 102 may be designated as the default data cache that is used to store data associated with execution entities that have not specifically been assigned to one of the other P-caches in the L1 cache. For example, such execution entities may be assigned a default EID 0 by the operating system. The P1- and P2-caches entities having EIDs 1 and 2, respectively. In one embodiment, the P0-cache may be a larger memory than either the P1- or P2-cache since it is the default data cache. - Other components of the
CPU 200 according to one example configuration are illustrated in FIG. 4. A bus front unit (BFU) 404 forms the interface to the front side orhost bus 203. TheBFU 404 may include address drivers and receivers, write buffers, data bus transceivers, bus control logic, bus master control, and parity generation and control. - The instruction path is first described below. Instructions retrieved by the
BFU 404 from either themain memory 206 or from theL2 cache 204 may be stored in aninstruction cache 406 that is part of theL1 cache 202. Theinternal instruction cache 406 may keep copies of the most frequently used instructions. According to some embodiments, instructions are fetched along with EIDs from either themain memory 206 or theL2 cache 204 and stored in theinstruction cache 406. An instruction buffer and decodelogic 408 decodes a selected instruction and associated EID from theinstruction cache 406 and produces one or more micro-operations along with corresponding EIDs. - In an alternative embodiment, the instructions are stored in different instruction memory regions of the
memory 206 according to different EIDs. However, in this embodiment, the EIDs are not stored along with the instructions. When theCPU 200 fetches an instruction, an associated EID is not retrieved. Instead, theCPU 200 determines the EID of the fetched instruction based on the address location where the instruction is stored. This may be performed, for example, by thedecode logic 408. Thus, the EID of an instruction is determined based on which instruction memory region the instruction is fetched from. Once the EID is determined by theCPU 200, it can be attached to the decoded micro-operations and stored in theinstruction queue 412. - In yet a further embodiment in which EIDs are not stored along with instructions in memory, multiple program counters and microsequencers may be included in the
CPU 200 that are assigned to corresponding threads. This embodiment is described further below. - The output port of the instruction buffer and decode
logic 408 may be coupled to aninstruction queue 412, which stores the micro-operations along with associated EIDs. The output port of theinstruction queue 412 is routed to asequencer 414. Thesequencer 414 may include multiplemicrosequencer units microsequencer unit 430 may be configured to process micro-operations associated with EID 0, themicrosequencer 432 may be configured to process micro-operations associated withEID 1, and themicrosequencer 434 may be configured to process micro-operations associated with EID 2. The micro-operations processed by themicrosequencers instruction queue 412. According to one embodiment, themicrosequencers microsequencers control logic 436 in thesequencer 414. - In one embodiment, the EIDs associated with micro-operations are originally retrieved from memory along with instructions of the different execution entities. In a further embodiment in which instructions are are not stored along with instructions, each microsequencer may be independently configured to fetch instructions of corresponding execution entities. Thus, a first microsequencer fetches instructions associated with a first execution entity, a second microsequencer fetches instructions associated with a second execution entity, and so forth. The EIDs of a fetched instruction may thus be determined based on which of the microsequencers fetched the instruction.
- An execution entity typically includes a number of instructions that are executed in some program order. By default, instruction addresses are simply incremented to fetch the next instruction. If a jump or other conditional branch occurs, then a target address is specified for the address of the next instruction. Thus, the address of the memory location where the next instruction is stored is known. A program counter may be used to keep track of the program order of instructions. A microsequencer works in conjunction with the program counter to execute instructions. To fetch an instruction, the microsequencer may ask a fetch unit, located in the
BFU 404 for example, to fetch an instruction with an address stored in the program counter. Thus, fetched instructions may be identified as belonging to an execution entity since the microsequencer already knows (from the program counter) the address of the next instruction. - Thus, for example, given a system with several threads, two or more independent program counters may be used. For example, three program counters PC0, PC1, and PC2 may be associated with
microsequencers microsequencer 430 keeps track of the program sequence for a first thread, the combination of PC1 andmicrosequencer 432 keeps track of the program sequence for a second thread, and so forth. When an instruction pointed to by PC0 is fetched, theCPU 200 knows that the instruction belongs to a first thread having, for example, EID 0. The EID is then attached to the instruction in theCPU 200 and to subsequently decoded micro-operations that are stored in theinstruction queue 412 for execution by themicrosequencer sequencer 414. - As shown in FIG. 4, the output port of the
sequencer 414 is provided to a pipeline back-end block 415 that includes various functional units, such as for example, an earlybranch execution unit 416, afast decoder unit 418, an arithmetic/logic unit (ALU) 420, and an address generator unit (AGU) 422. During execution of one or more micro-operations by thesequencer 414, these functional units may be accessed to perform requested functions. - The pipeline back-
end block 415 also includes register files 424, 426, and 428. The register files 424, 426 and 428 in theCPU 200 correspond to the three EID groups EID 0,EID 1, and EID 2. The register files 424, 426 and 428 may each include control registers, status registers, flag registers, and general purpose registers. The register files 424, 426, and 428 are updated by the functional units in the pipeline back-end block 415 during operation. According to one embodiment, the register files 424, 426, and 428 may also be accessible independently and concurrently. - In the illustrated embodiment, requests associated with different EIDs may be processed concurrently provided that there are no dependencies among the requests, and further, the multiple requests do not need to utilize the same
functional units microsequencers - In the data path of the
CPU 200, a store buffer 450 (for write operations) and a load buffer 452 (for read operations) store data that are retrieved from or targeted for theBFU 404. The store andload buffers internal data bus 454 that is coupled to several units, including the P0-cache 102, the P1-cache 104, the P2-cache 106, the pipeline back-end block 415, and a translation look aside buffer (TLB) 456. - Addresses of instructions in the
instruction cache 406 are fed to theTLB 456, which is basically a high speed memory in theCPU 200 that translates the virtual address from theinstruction cache 406 into a physical address to access thedata cache modules - Based on the multi-unit data cache availability, the
control logic 436 in themicrocode sequencer 414 may select an appropriate instruction for processing by one of themicrosequencers microsequencers CPU 200 with concurrent access to data in the L1multi-unit cache 202. - In some embodiments, the
control logic 436 of thesequencer 414 may also consider possible load/store ordering, outstanding data cache refilling, and other issues. For example, in one embodiment, instructions associated with a request that has been determined to have a high hit ratio may be scheduled first, as may instructions of a real-time constrained execution entity having high priority. - Referring to FIG. 5, the general flow of an instruction execution sequence according to an embodiment is illustrated. Instructions are fetched from
main memory 206 or L2 cache 204 (at 302) by theCPU 200 over thehost bus 203. In one embodiment, associated EIDs are retrieved with the instructions. In another embodiment, the associated EIDs are not stored and thus are not retrieved. The fetched instructions are then translated into internal micro-operations (at 304) by thedecoder stage 408, with a corresponding EID attached to each micro-operation. The EID may be the one fetched with the instruction or it may be determined by theCPU 200 based on the address location of the instruction or which microsequencer fetched the instruction. Next, the translated micro-operation is stored in the instruction queue 412 (at 306). The micro-operation is then delivered to one of themicrosequencers cache controller 108 and an appropriate request is sent to a corresponding P-cache (102, 104, or 106). The data access request is then completed in the assigned P-cache (at 312). - While the embodiments described include a multi-unit cache memory to store data, it is contemplated that the multi-unit cache memory may be adapted to store instructions of different execution entities in further embodiments. In such embodiments, the information stored in the multi-unit cache memory includes the instructions themselves.
- While the invention has been disclosed with respect to a limited number of embodiments, those skilled in the art will appreciate numerous modifications and variations therefrom. It is intended that the appended claims cover all such modifications and variations as fall within the true spirit and scope of the invention.
Claims (29)
1. A system comprising:
a processor;
a plurality of execution entities executable on the processor; and
a buffer memory having multiple buffer sections, each buffer section adapted to store information associated with requests from a corresponding one of the multiple execution entities.
2. The system of , further comprising a software layer adapted to assign identifiers to the execution identifiers and to assign each buffer section an identifier.
claim 1
3. The system of , further comprising a controller operatively coupled to the buffer memory to select one of the buffer sections based on an identifier associated with a request from an execution entity.
claim 2
4. The system of , wherein the execution entities include processes.
claim 1
5. The system of , wherein the execution entities include threads.
claim 1
6. The system of , wherein the buffer memory includes a cache memory having multiple cache sections.
claim 1
7. The system of , wherein the cache sections are configurable to have different attributes.
claim 6
8. The system of , wherein the attributes include cache line replacement policies.
claim 7
9. The system of , wherein the attributes include cache update policies.
claim 7
10. The system of , wherein the attributes include cache organization.
claim 7
11. The system of , wherein the cache memory includes a multi-level cache memory in which at least one level includes a multi-unit cache memory having multiple cache sections.
claim 6
12. The system of , wherein the execution entities process data according to different temporal and spatial locality characteristics, and each buffer section is configured based on the temporal and spatial locality characteristics.
claim 1
13. A cache memory for use in a system having multiple execution entities, the cache memory comprising:
a cache controller; and
a plurality of cache sections, the cache controller adapted to store information in one of the cache sections based on which execution entity the information is associated with.
14. The cache memory of , wherein the cache controller includes storage elements programmable with identifiers to identify the one or more execution entities that each cache section is associated with.
claim 13
15. A method of storing data in a memory having multiple sections, the memory located in a system having multiple execution entities, the method comprising:
assigning an identifier to each execution entity;
retrieving an instruction of one of the execution entities;
decoding an identifier associated with the instruction; and
storing information associated with the instruction in one of the memory sections based on the identifier.
16. The method of wherein decoding includes retrieving the identifier from a storage location.
claim 15
17. A method of setting up a cache memory having multiple cache sections, the cache memory located in a system having multiple execution entities, the method comprising:
assigning an identifier to each execution entity based on which of the multiple cache sections is to be used for the execution entity; and
configuring each of the cache sections based on the cache usage characteristics of the one or more execution entities assigned to the cache section.
18. The method of , wherein the configuring includes setting attributes of each cache section.
claim 17
19. The method of , wherein the attribute setting includes setting a replacement policy for each cache section.
claim 18
20. The method of , wherein the attribute setting includes setting an update policy for each cache section.
claim 18
21. A memory subsystem in a system having multiple execution entities that generate instructions, comprising:
a controller; and
a multi-unit buffer memory having multiple buffers,
the controller adapted to select one of the buffers to store information associated with an instruction based on which execution entity generated the instruction.
22. The memory subsystem of , wherein the controller includes storage elements corresponding to each of the buffers that are programmable to values identifying the execution entities.
claim 21
23. An article including a storage medium containing instructions for managing memory in a system, the system having a processor, a memory with multiple memory sections, and multiple execution entities executable on the processor, the instructions when executed causing the system to:
assign each memory section to correspond to one or more of the execution entities; and
configure attributes of each of the memory sections based on the memory usage characteristics of the one or more execution entities assigned to the memory section.
24. A processor located in a system having multiple execution entities, comprising:
a cache memory having multiple cache sections each assigned to an execution entity; and
a sequencer having multiple segments each assigned to an execution entity,
the sequencer adapted to receive instructions from multiple execution entities, each segment of the sequencer capable of executing the received instructions concurrently and accessing the cache sections concurrently during execution.
25. A system comprising:
program execution entities associated with identifiers; and
a multi-unit cache memory having multiple cache sections adapted to store information associated with requests from the execution entities, each cache section storing the information based on an identifier.
26. The system of further comprising a memory in which requests and associated identifiers are stored.
claim 25
27. The system of , further comprising a processor and a memory to store instructions associated with requests from the execution entities, the processor adapted to retrieve an instruction from the memory and to determine an identifier associated with the instruction based on an address location of the instruction in the memory.
claim 25
28. The system of , wherein the requests include instructions, the system further comprising a processor having a plurality of program counters and corresponding microsequencers, each microsequencer adapted to fetch instructions associated with a corresponding execution entity based on an address contained in the program counter.
claim 25
29. The system of , wherein the identifier of an instruction is determined based on which microsequencer fetched the instruction.
claim 28
Priority Applications (8)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US09/207,278 US6349363B2 (en) | 1998-12-08 | 1998-12-08 | Multi-section cache with different attributes for each section |
AU64971/99A AU6497199A (en) | 1998-12-08 | 1999-09-22 | Buffer memory management in a system having multiple execution entities |
CNB998160911A CN1179276C (en) | 1998-12-08 | 1999-09-22 | Buffer memory management in system haivng multiple execution entities |
GB0111896A GB2359163B (en) | 1998-12-08 | 1999-09-22 | Buffer memory management in a system having multiple execution entities |
DE19983793T DE19983793B4 (en) | 1998-12-08 | 1999-09-22 | A system comprising a processor on which a plurality of concurrent execution entities are executed, and a cache memory having multiple cache sections associated with execution entities |
PCT/US1999/021249 WO2000034870A1 (en) | 1998-12-08 | 1999-09-22 | Buffer memory management in a system having multiple execution entities |
HK01107440A HK1036662A1 (en) | 1998-12-08 | 2001-10-24 | Buffer memory management in a system having multiple execution entities |
US10/007,594 US6470422B2 (en) | 1998-12-08 | 2001-11-09 | Buffer memory management in a system having multiple execution entities |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US09/207,278 US6349363B2 (en) | 1998-12-08 | 1998-12-08 | Multi-section cache with different attributes for each section |
Related Child Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US10/007,594 Division US6470422B2 (en) | 1998-12-08 | 2001-11-09 | Buffer memory management in a system having multiple execution entities |
Publications (2)
Publication Number | Publication Date |
---|---|
US20010049770A1 true US20010049770A1 (en) | 2001-12-06 |
US6349363B2 US6349363B2 (en) | 2002-02-19 |
Family
ID=22769873
Family Applications (2)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US09/207,278 Expired - Lifetime US6349363B2 (en) | 1998-12-08 | 1998-12-08 | Multi-section cache with different attributes for each section |
US10/007,594 Expired - Lifetime US6470422B2 (en) | 1998-12-08 | 2001-11-09 | Buffer memory management in a system having multiple execution entities |
Family Applications After (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US10/007,594 Expired - Lifetime US6470422B2 (en) | 1998-12-08 | 2001-11-09 | Buffer memory management in a system having multiple execution entities |
Country Status (7)
Country | Link |
---|---|
US (2) | US6349363B2 (en) |
CN (1) | CN1179276C (en) |
AU (1) | AU6497199A (en) |
DE (1) | DE19983793B4 (en) |
GB (1) | GB2359163B (en) |
HK (1) | HK1036662A1 (en) |
WO (1) | WO2000034870A1 (en) |
Cited By (32)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20030061444A1 (en) * | 2001-09-14 | 2003-03-27 | Seagate Technology Llc | Method and system for cache management algorithm selection |
US6675261B2 (en) * | 2000-12-22 | 2004-01-06 | Oblix, Inc. | Request based caching of data store data |
US20040015660A1 (en) * | 2002-07-22 | 2004-01-22 | Caroline Benveniste | Cache configuration for compressed memory systems |
US6760743B1 (en) * | 2000-01-04 | 2004-07-06 | International Business Machines Corporation | Instruction memory system for multi-processor environment and disjoint tasks |
US6772276B2 (en) * | 2002-01-04 | 2004-08-03 | Intel Corporation | Flash memory command abstraction |
US6832378B1 (en) * | 2000-06-20 | 2004-12-14 | International Business Machines Corporation | Parallel software processing system |
US20050050455A1 (en) * | 2003-08-29 | 2005-03-03 | Yee Sunny K. | Method and apparatus for supporting object caching in a web presentation architecture |
US20050076291A1 (en) * | 2003-10-01 | 2005-04-07 | Yee Sunny K. | Method and apparatus for supporting page localization management in a Web presentation architecture |
US20060123395A1 (en) * | 2004-12-07 | 2006-06-08 | International Business Machines Corporation | Controlling user intervention in a multi-processing computer system |
US20070136530A1 (en) * | 2003-11-12 | 2007-06-14 | Tetsuya Tanaka | Cache memory and control method thereof |
US20090164698A1 (en) * | 2007-12-24 | 2009-06-25 | Yung-Li Ji | Nonvolatile storage device with NCQ supported and writing method for a nonvolatile storage device |
US20090204740A1 (en) * | 2004-10-25 | 2009-08-13 | Robert Bosch Gmbh | Method and Device for Performing Switchover Operations in a Computer System Having at Least Two Execution Units |
US7673047B2 (en) | 2000-12-22 | 2010-03-02 | Oracle International Corporation | Determining a user's groups |
US7711818B2 (en) | 2000-12-22 | 2010-05-04 | Oracle International Corporation | Support for multiple data stores |
US7765298B2 (en) | 2001-11-30 | 2010-07-27 | Oracle International Corporation | Impersonation in an access system |
US7802174B2 (en) | 2000-12-22 | 2010-09-21 | Oracle International Corporation | Domain based workflows |
US7840658B2 (en) | 2002-05-15 | 2010-11-23 | Oracle International Corporation | Employing job code attributes in provisioning |
US7882132B2 (en) | 2003-10-09 | 2011-02-01 | Oracle International Corporation | Support for RDBMS in LDAP system |
US7904487B2 (en) | 2003-10-09 | 2011-03-08 | Oracle International Corporation | Translating data access requests |
US7937655B2 (en) | 2000-12-22 | 2011-05-03 | Oracle International Corporation | Workflows with associated processes |
US7949833B1 (en) | 2005-01-13 | 2011-05-24 | Marvell International Ltd. | Transparent level 2 cache controller |
US8015600B2 (en) | 2000-12-22 | 2011-09-06 | Oracle International Corporation | Employing electronic certificate workflows |
KR101136610B1 (en) | 2005-04-05 | 2012-04-23 | 인텔 코오퍼레이션 | Sequencer address management |
US8347034B1 (en) * | 2005-01-13 | 2013-01-01 | Marvell International Ltd. | Transparent level 2 cache that uses independent tag and valid random access memory arrays for cache access |
US20130159631A1 (en) * | 2011-12-19 | 2013-06-20 | Thomas Seufert | Transactional-consistent cache for database objects |
US20140025868A1 (en) * | 2008-10-17 | 2014-01-23 | Seagate Technology Llc | System and method for managing storage device caching |
US20140095716A1 (en) * | 2012-09-28 | 2014-04-03 | International Business Machines Corporation | Maximizing resources in a multi-application processing environement |
CN105094686A (en) * | 2014-05-09 | 2015-11-25 | 华为技术有限公司 | Data caching method, cache and computer system |
US20160335021A1 (en) * | 2015-05-11 | 2016-11-17 | Mentor Graphics Corporation | Memory corruption protection by tracing memory |
US20180292991A1 (en) * | 2017-04-11 | 2018-10-11 | Micron Technology, Inc. | Memory protocol with programmable buffer and cache size |
US10534540B2 (en) | 2016-06-06 | 2020-01-14 | Micron Technology, Inc. | Memory protocol |
US10678441B2 (en) | 2016-05-05 | 2020-06-09 | Micron Technology, Inc. | Non-deterministic memory protocol |
Families Citing this family (116)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP3133004B2 (en) * | 1996-11-21 | 2001-02-05 | 株式会社日立製作所 | Disk array device and control method therefor |
JP2000010860A (en) * | 1998-06-16 | 2000-01-14 | Hitachi Ltd | Cache memory control circuit, processor, processor system, and parallel processor system |
US6378066B1 (en) * | 1999-02-04 | 2002-04-23 | Sun Microsystems, Inc. | Method, apparatus, and article of manufacture for developing and executing data flow programs, and optimizing user input specifications |
EP1098249A1 (en) * | 1999-11-05 | 2001-05-09 | Emc Corporation | Segmenting cache to provide varying service levels |
US6457102B1 (en) * | 1999-11-05 | 2002-09-24 | Emc Corporation | Cache using multiple LRU's |
US6791555B1 (en) * | 2000-06-23 | 2004-09-14 | Micron Technology, Inc. | Apparatus and method for distributed memory control in a graphics processing system |
US6889291B1 (en) * | 2000-06-30 | 2005-05-03 | Intel Corporation | Method and apparatus for cache replacement for a multiple variable-way associative cache |
AU2001286410A1 (en) * | 2000-07-31 | 2002-02-13 | Morphics Technology, Inc. | Method and apparatus for time-sliced and multi-threaded data processing in a communication system |
ATE548695T1 (en) | 2000-08-21 | 2012-03-15 | Texas Instruments France | SOFTWARE CONTROLLED CACHE MEMORY CONFIGURATION |
EP1215583A1 (en) * | 2000-12-15 | 2002-06-19 | Texas Instruments Incorporated | Cache with tag entries having additional qualifier fields |
US6684298B1 (en) | 2000-11-09 | 2004-01-27 | University Of Rochester | Dynamic reconfigurable memory hierarchy |
US6701417B2 (en) * | 2001-04-11 | 2004-03-02 | Sun Microsystems, Inc. | Method and apparatus for supporting multiple cache line invalidations per cycle |
US6684297B2 (en) * | 2001-04-11 | 2004-01-27 | Sun Microsystems, Inc. | Reverse directory for facilitating accesses involving a lower-level cache |
US6965970B2 (en) * | 2001-09-27 | 2005-11-15 | Intel Corporation | List based method and apparatus for selective and rapid cache flushes |
US6857937B2 (en) * | 2002-05-30 | 2005-02-22 | Komag, Inc. | Lapping a head while powered up to eliminate expansion of the head due to heating |
US7133972B2 (en) | 2002-06-07 | 2006-11-07 | Micron Technology, Inc. | Memory hub with internal cache and/or memory access prediction |
US7200024B2 (en) * | 2002-08-02 | 2007-04-03 | Micron Technology, Inc. | System and method for optically interconnecting memory devices |
US7117316B2 (en) * | 2002-08-05 | 2006-10-03 | Micron Technology, Inc. | Memory hub and access method having internal row caching |
US7254331B2 (en) * | 2002-08-09 | 2007-08-07 | Micron Technology, Inc. | System and method for multiple bit optical data transmission in memory systems |
US7149874B2 (en) * | 2002-08-16 | 2006-12-12 | Micron Technology, Inc. | Memory hub bypass circuit and method |
US6754117B2 (en) | 2002-08-16 | 2004-06-22 | Micron Technology, Inc. | System and method for self-testing and repair of memory modules |
US6820181B2 (en) | 2002-08-29 | 2004-11-16 | Micron Technology, Inc. | Method and system for controlling memory accesses to memory modules having a memory hub architecture |
US7836252B2 (en) * | 2002-08-29 | 2010-11-16 | Micron Technology, Inc. | System and method for optimizing interconnections of memory devices in a multichip module |
US7102907B2 (en) * | 2002-09-09 | 2006-09-05 | Micron Technology, Inc. | Wavelength division multiplexed memory module, memory system and method |
US7076609B2 (en) * | 2002-09-20 | 2006-07-11 | Intel Corporation | Cache sharing for a chip multiprocessor or multiprocessing system |
US6711635B1 (en) * | 2002-09-30 | 2004-03-23 | Western Digital Technologies, Inc. | Disk drive employing thresholds for cache memory allocation |
US6823430B2 (en) * | 2002-10-10 | 2004-11-23 | International Business Machines Corporation | Directoryless L0 cache for stall reduction |
DE602004023372D1 (en) * | 2003-02-24 | 2009-11-12 | Nxp Bv | CACHE STORE TRASHING REDUCTION OF PARTICULAR CODE PIECES |
US7190367B2 (en) * | 2003-03-25 | 2007-03-13 | Mitsubishi Electric Research Laboratories, Inc. | Method, apparatus, and system for rendering using a progressive cache |
US7245145B2 (en) * | 2003-06-11 | 2007-07-17 | Micron Technology, Inc. | Memory module and method having improved signal routing topology |
US7120727B2 (en) | 2003-06-19 | 2006-10-10 | Micron Technology, Inc. | Reconfigurable memory module and method |
US7107415B2 (en) * | 2003-06-20 | 2006-09-12 | Micron Technology, Inc. | Posted write buffers and methods of posting write requests in memory modules |
US7428644B2 (en) * | 2003-06-20 | 2008-09-23 | Micron Technology, Inc. | System and method for selective memory module power management |
US7260685B2 (en) | 2003-06-20 | 2007-08-21 | Micron Technology, Inc. | Memory hub and access method having internal prefetch buffers |
US7389364B2 (en) * | 2003-07-22 | 2008-06-17 | Micron Technology, Inc. | Apparatus and method for direct memory access in a hub-based memory system |
US7210059B2 (en) * | 2003-08-19 | 2007-04-24 | Micron Technology, Inc. | System and method for on-board diagnostics of memory modules |
US7133991B2 (en) * | 2003-08-20 | 2006-11-07 | Micron Technology, Inc. | Method and system for capturing and bypassing memory transactions in a hub-based memory system |
US7136958B2 (en) | 2003-08-28 | 2006-11-14 | Micron Technology, Inc. | Multiple processor system and method including multiple memory hub modules |
US20050050237A1 (en) * | 2003-08-28 | 2005-03-03 | Jeddeloh Joseph M. | Memory module and method having on-board data search capabilities and processor-based system using such memory modules |
US7310752B2 (en) | 2003-09-12 | 2007-12-18 | Micron Technology, Inc. | System and method for on-board timing margin testing of memory modules |
US7194593B2 (en) * | 2003-09-18 | 2007-03-20 | Micron Technology, Inc. | Memory hub with integrated non-volatile memory |
US7120743B2 (en) * | 2003-10-20 | 2006-10-10 | Micron Technology, Inc. | Arbitration system and method for memory responses in a hub-based memory system |
US7234070B2 (en) * | 2003-10-27 | 2007-06-19 | Micron Technology, Inc. | System and method for using a learning sequence to establish communications on a high-speed nonsynchronous interface in the absence of clock forwarding |
US7330992B2 (en) | 2003-12-29 | 2008-02-12 | Micron Technology, Inc. | System and method for read synchronization of memory modules |
US7216196B2 (en) * | 2003-12-29 | 2007-05-08 | Micron Technology, Inc. | Memory hub and method for memory system performance monitoring |
US7173863B2 (en) * | 2004-03-08 | 2007-02-06 | Sandisk Corporation | Flash controller cache architecture |
US7213107B2 (en) * | 2003-12-31 | 2007-05-01 | Intel Corporation | Dedicated cache memory |
US7188219B2 (en) * | 2004-01-30 | 2007-03-06 | Micron Technology, Inc. | Buffer control system and method for a memory system having outstanding read and write request buffers |
US7356651B2 (en) * | 2004-01-30 | 2008-04-08 | Piurata Technologies, Llc | Data-aware cache state machine |
US7181584B2 (en) * | 2004-02-05 | 2007-02-20 | Micron Technology, Inc. | Dynamic command and/or address mirroring system and method for memory modules |
US7412574B2 (en) * | 2004-02-05 | 2008-08-12 | Micron Technology, Inc. | System and method for arbitration of memory responses in a hub-based memory system |
US7788451B2 (en) | 2004-02-05 | 2010-08-31 | Micron Technology, Inc. | Apparatus and method for data bypass for a bi-directional data bus in a hub-based memory sub-system |
US7797492B2 (en) * | 2004-02-20 | 2010-09-14 | Anoop Mukker | Method and apparatus for dedicating cache entries to certain streams for performance optimization |
US7366864B2 (en) * | 2004-03-08 | 2008-04-29 | Micron Technology, Inc. | Memory hub architecture having programmable lane widths |
US20050206648A1 (en) * | 2004-03-16 | 2005-09-22 | Perry Ronald N | Pipeline and cache for processing data progressively |
US7257683B2 (en) * | 2004-03-24 | 2007-08-14 | Micron Technology, Inc. | Memory arbitration system and method having an arbitration packet protocol |
US7120723B2 (en) * | 2004-03-25 | 2006-10-10 | Micron Technology, Inc. | System and method for memory hub-based expansion bus |
US7213082B2 (en) * | 2004-03-29 | 2007-05-01 | Micron Technology, Inc. | Memory hub and method for providing memory sequencing hints |
US7447240B2 (en) * | 2004-03-29 | 2008-11-04 | Micron Technology, Inc. | Method and system for synchronizing communications links in a hub-based memory system |
US6980042B2 (en) * | 2004-04-05 | 2005-12-27 | Micron Technology, Inc. | Delay line synchronizer apparatus and method |
US7590797B2 (en) | 2004-04-08 | 2009-09-15 | Micron Technology, Inc. | System and method for optimizing interconnections of components in a multichip memory module |
US7162567B2 (en) * | 2004-05-14 | 2007-01-09 | Micron Technology, Inc. | Memory hub and method for memory sequencing |
US7222213B2 (en) * | 2004-05-17 | 2007-05-22 | Micron Technology, Inc. | System and method for communicating the synchronization status of memory modules during initialization of the memory modules |
US7363419B2 (en) * | 2004-05-28 | 2008-04-22 | Micron Technology, Inc. | Method and system for terminating write commands in a hub-based memory system |
US7519788B2 (en) * | 2004-06-04 | 2009-04-14 | Micron Technology, Inc. | System and method for an asynchronous data buffer having buffer write and read pointers |
US7310748B2 (en) | 2004-06-04 | 2007-12-18 | Micron Technology, Inc. | Memory hub tester interface and method for use thereof |
US7685354B1 (en) * | 2004-06-30 | 2010-03-23 | Sun Microsystems, Inc. | Multiple-core processor with flexible mapping of processor cores to cache banks |
US7873776B2 (en) * | 2004-06-30 | 2011-01-18 | Oracle America, Inc. | Multiple-core processor with support for multiple virtual processors |
US7392331B2 (en) * | 2004-08-31 | 2008-06-24 | Micron Technology, Inc. | System and method for transmitting data packets in a computer system having a memory hub architecture |
US7590803B2 (en) * | 2004-09-23 | 2009-09-15 | Sap Ag | Cache eviction |
US20060074872A1 (en) * | 2004-09-30 | 2006-04-06 | International Business Machines Corporation | Adaptive database buffer memory management using dynamic SQL statement cache statistics |
US7472224B1 (en) | 2004-10-01 | 2008-12-30 | Advanced Micro Devices, Inc. | Reconfigurable processing node including first and second processor cores |
US7257678B2 (en) * | 2004-10-01 | 2007-08-14 | Advanced Micro Devices, Inc. | Dynamic reconfiguration of cache memory |
US7580915B2 (en) * | 2004-12-14 | 2009-08-25 | Sap Ag | Socket-like communication API for C |
US7600217B2 (en) * | 2004-12-14 | 2009-10-06 | Sap Ag | Socket-like communication API for Java |
US7882299B2 (en) * | 2004-12-21 | 2011-02-01 | Sandisk Corporation | System and method for use of on-chip non-volatile memory write cache |
US20060143398A1 (en) * | 2004-12-23 | 2006-06-29 | Stefan Rau | Method and apparatus for least recently used (LRU) software cache |
US20060143256A1 (en) | 2004-12-28 | 2006-06-29 | Galin Galchev | Cache region concept |
US20060143389A1 (en) * | 2004-12-28 | 2006-06-29 | Frank Kilian | Main concept for common cache management |
US7451275B2 (en) * | 2004-12-28 | 2008-11-11 | Sap Ag | Programming models for storage plug-ins |
US7512737B2 (en) * | 2004-12-28 | 2009-03-31 | Sap Ag | Size based eviction implementation |
US7971001B2 (en) * | 2004-12-28 | 2011-06-28 | Sap Ag | Least recently used eviction implementation |
US8204931B2 (en) | 2004-12-28 | 2012-06-19 | Sap Ag | Session management within a multi-tiered enterprise network |
US7694065B2 (en) | 2004-12-28 | 2010-04-06 | Sap Ag | Distributed cache architecture |
US7552153B2 (en) * | 2004-12-28 | 2009-06-23 | Sap Ag | Virtual machine monitoring using shared memory |
US7437516B2 (en) * | 2004-12-28 | 2008-10-14 | Sap Ag | Programming models for eviction policies |
US7552284B2 (en) * | 2004-12-28 | 2009-06-23 | Sap Ag | Least frequently used eviction implementation |
US7539821B2 (en) * | 2004-12-28 | 2009-05-26 | Sap Ag | First in first out eviction implementation |
US7810083B2 (en) * | 2004-12-30 | 2010-10-05 | Intel Corporation | Mechanism to emulate user-level multithreading on an OS-sequestered sequencer |
US20060168407A1 (en) * | 2005-01-26 | 2006-07-27 | Micron Technology, Inc. | Memory hub system and method having large virtual page size |
US7516277B2 (en) * | 2005-04-28 | 2009-04-07 | Sap Ag | Cache monitoring using shared memory |
US7831634B2 (en) | 2005-04-29 | 2010-11-09 | Sap Ag | Initializing a cache region using a generated cache region configuration structure |
US7581066B2 (en) * | 2005-04-29 | 2009-08-25 | Sap Ag | Cache isolation model |
US7966412B2 (en) * | 2005-07-19 | 2011-06-21 | Sap Ag | System and method for a pluggable protocol handler |
US7788243B2 (en) * | 2006-09-08 | 2010-08-31 | Sybase, Inc. | System and methods for optimizing data transfer among various resources in a distributed environment |
US8136119B2 (en) * | 2006-11-30 | 2012-03-13 | Dell Products L.P. | Method, apparatus and media for managing information model jobs |
US7831772B2 (en) * | 2006-12-12 | 2010-11-09 | Sybase, Inc. | System and methodology providing multiple heterogeneous buffer caches |
US8074274B2 (en) * | 2006-12-29 | 2011-12-06 | Intel Corporation | User-level privilege management |
US20080163063A1 (en) * | 2006-12-29 | 2008-07-03 | Sap Ag | Graphical user interface system and method for presenting information related to session and cache objects |
US8542323B2 (en) * | 2007-12-21 | 2013-09-24 | Sony Corporation | Touch sensitive wireless navigation device for remote control |
US7870339B2 (en) * | 2008-01-11 | 2011-01-11 | International Business Machines Corporation | Extract cache attribute facility and instruction therefore |
DE112008003788T5 (en) * | 2008-03-27 | 2011-02-24 | Hewlett-Packard Co. Intellectual Property Administration, Fort Collins | Raid array access by an operating system that does not know the RAID array |
US8166229B2 (en) | 2008-06-30 | 2012-04-24 | Intel Corporation | Apparatus and method for multi-level cache utilization |
CN101515295B (en) * | 2009-03-23 | 2010-12-01 | 浙江大学 | Realization method for supporting high-speed buffer of hardware database on chip |
JP5785357B2 (en) | 2009-06-25 | 2015-09-30 | スパンション エルエルシー | Computer system provided with reconfiguration arithmetic device and reconfiguration arithmetic device |
US8825685B2 (en) | 2009-11-16 | 2014-09-02 | Symantec Corporation | Selective file system caching based upon a configurable cache map |
EP2388704A1 (en) * | 2010-05-17 | 2011-11-23 | Thomson Licensing | Method of optimization of cache memory management and corresponding apparatus |
US8589627B2 (en) * | 2010-08-27 | 2013-11-19 | Advanced Micro Devices, Inc. | Partially sectored cache |
US8868843B2 (en) | 2011-11-30 | 2014-10-21 | Advanced Micro Devices, Inc. | Hardware filter for tracking block presence in large caches |
US9223709B1 (en) | 2012-03-06 | 2015-12-29 | Marvell International Ltd. | Thread-aware cache memory management |
US9235513B2 (en) * | 2012-10-18 | 2016-01-12 | International Business Machines Corporation | Cache management based on physical memory device characteristics |
US9251070B2 (en) * | 2012-12-19 | 2016-02-02 | Advanced Micro Devices, Inc. | Methods and apparatus for multi-level cache hierarchies |
CN104424119B (en) * | 2013-08-26 | 2018-07-06 | 联想(北京)有限公司 | Memory space configuration method and device |
TWI584121B (en) * | 2015-08-19 | 2017-05-21 | 旺宏電子股份有限公司 | Buffer cache device metod for managing the same and applying system thereof |
US20170083441A1 (en) * | 2015-09-23 | 2017-03-23 | Qualcomm Incorporated | Region-based cache management |
CN106649139B (en) * | 2016-12-29 | 2020-01-10 | 北京奇虎科技有限公司 | Data elimination method and device based on multiple caches |
Family Cites Families (19)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US4928239A (en) * | 1986-06-27 | 1990-05-22 | Hewlett-Packard Company | Cache memory with variable fetch and replacement schemes |
US4905141A (en) * | 1988-10-25 | 1990-02-27 | International Business Machines Corporation | Partitioned cache memory with partition look-aside table (PLAT) for early partition assignment identification |
US5875464A (en) * | 1991-12-10 | 1999-02-23 | International Business Machines Corporation | Computer system with private and shared partitions in cache |
US5479636A (en) | 1992-11-16 | 1995-12-26 | Intel Corporation | Concurrent cache line replacement method and apparatus in microprocessor system with write-back cache memory |
JP2583010B2 (en) * | 1993-01-07 | 1997-02-19 | インターナショナル・ビジネス・マシーンズ・コーポレイション | Method of maintaining consistency between local index table and global index table in multi-tier index structure |
JPH06282488A (en) * | 1993-03-25 | 1994-10-07 | Mitsubishi Electric Corp | Cache storage device |
US5809524A (en) | 1994-01-04 | 1998-09-15 | Intel Corporation | Method and apparatus for cache memory replacement line identification |
US5680573A (en) * | 1994-07-12 | 1997-10-21 | Sybase, Inc. | Method of buffering data objects in a database |
US5960194A (en) * | 1995-09-11 | 1999-09-28 | International Business Machines Corporation | Method for generating a multi-tiered index for partitioned data |
US5701432A (en) * | 1995-10-13 | 1997-12-23 | Sun Microsystems, Inc. | Multi-threaded processing system having a cache that is commonly accessible to each thread |
GB2311880A (en) * | 1996-04-03 | 1997-10-08 | Advanced Risc Mach Ltd | Partitioned cache memory |
EP0856798B1 (en) | 1997-01-30 | 2004-09-29 | STMicroelectronics Limited | A cache system |
US5963972A (en) * | 1997-02-24 | 1999-10-05 | Digital Equipment Corporation | Memory architecture dependent program mapping |
US6058456A (en) * | 1997-04-14 | 2000-05-02 | International Business Machines Corporation | Software-managed programmable unified/split caching mechanism for instructions and data |
US6092149A (en) * | 1997-05-28 | 2000-07-18 | Western Digital Corporation | Disk drive cache system using a dynamic priority sequential stream of data segments continuously adapted according to prefetched sequential random, and repeating types of accesses |
US6161166A (en) * | 1997-11-10 | 2000-12-12 | International Business Machines Corporation | Instruction cache for multithreaded processor |
US6112280A (en) * | 1998-01-06 | 2000-08-29 | Hewlett-Packard Company | Method and apparatus for distinct instruction pointer storage in a partitioned cache memory |
US6205519B1 (en) * | 1998-05-27 | 2001-03-20 | Hewlett Packard Company | Cache management for a multi-threaded processor |
US6269425B1 (en) * | 1998-08-20 | 2001-07-31 | International Business Machines Corporation | Accessing data from a multiple entry fully associative cache buffer in a multithread data processing system |
-
1998
- 1998-12-08 US US09/207,278 patent/US6349363B2/en not_active Expired - Lifetime
-
1999
- 1999-09-22 DE DE19983793T patent/DE19983793B4/en not_active Expired - Fee Related
- 1999-09-22 AU AU64971/99A patent/AU6497199A/en not_active Abandoned
- 1999-09-22 WO PCT/US1999/021249 patent/WO2000034870A1/en active Application Filing
- 1999-09-22 CN CNB998160911A patent/CN1179276C/en not_active Expired - Fee Related
- 1999-09-22 GB GB0111896A patent/GB2359163B/en not_active Expired - Fee Related
-
2001
- 2001-10-24 HK HK01107440A patent/HK1036662A1/en not_active IP Right Cessation
- 2001-11-09 US US10/007,594 patent/US6470422B2/en not_active Expired - Lifetime
Cited By (49)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6760743B1 (en) * | 2000-01-04 | 2004-07-06 | International Business Machines Corporation | Instruction memory system for multi-processor environment and disjoint tasks |
US6832378B1 (en) * | 2000-06-20 | 2004-12-14 | International Business Machines Corporation | Parallel software processing system |
US7937655B2 (en) | 2000-12-22 | 2011-05-03 | Oracle International Corporation | Workflows with associated processes |
US6675261B2 (en) * | 2000-12-22 | 2004-01-06 | Oblix, Inc. | Request based caching of data store data |
US7673047B2 (en) | 2000-12-22 | 2010-03-02 | Oracle International Corporation | Determining a user's groups |
US9235649B2 (en) | 2000-12-22 | 2016-01-12 | Oracle International Corporation | Domain based workflows |
US8015600B2 (en) | 2000-12-22 | 2011-09-06 | Oracle International Corporation | Employing electronic certificate workflows |
US7802174B2 (en) | 2000-12-22 | 2010-09-21 | Oracle International Corporation | Domain based workflows |
US7711818B2 (en) | 2000-12-22 | 2010-05-04 | Oracle International Corporation | Support for multiple data stores |
US6883066B2 (en) * | 2001-09-14 | 2005-04-19 | Seagate Technology Llc | Method and system for cache management algorithm selection |
US20030061444A1 (en) * | 2001-09-14 | 2003-03-27 | Seagate Technology Llc | Method and system for cache management algorithm selection |
US7765298B2 (en) | 2001-11-30 | 2010-07-27 | Oracle International Corporation | Impersonation in an access system |
US6772276B2 (en) * | 2002-01-04 | 2004-08-03 | Intel Corporation | Flash memory command abstraction |
US7840658B2 (en) | 2002-05-15 | 2010-11-23 | Oracle International Corporation | Employing job code attributes in provisioning |
US7103722B2 (en) * | 2002-07-22 | 2006-09-05 | International Business Machines Corporation | Cache configuration for compressed memory systems |
US20040015660A1 (en) * | 2002-07-22 | 2004-01-22 | Caroline Benveniste | Cache configuration for compressed memory systems |
US20050050455A1 (en) * | 2003-08-29 | 2005-03-03 | Yee Sunny K. | Method and apparatus for supporting object caching in a web presentation architecture |
US20050076291A1 (en) * | 2003-10-01 | 2005-04-07 | Yee Sunny K. | Method and apparatus for supporting page localization management in a Web presentation architecture |
US7882132B2 (en) | 2003-10-09 | 2011-02-01 | Oracle International Corporation | Support for RDBMS in LDAP system |
US7904487B2 (en) | 2003-10-09 | 2011-03-08 | Oracle International Corporation | Translating data access requests |
US7502887B2 (en) | 2003-11-12 | 2009-03-10 | Panasonic Corporation | N-way set associative cache memory and control method thereof |
US20070136530A1 (en) * | 2003-11-12 | 2007-06-14 | Tetsuya Tanaka | Cache memory and control method thereof |
US20090204740A1 (en) * | 2004-10-25 | 2009-08-13 | Robert Bosch Gmbh | Method and Device for Performing Switchover Operations in a Computer System Having at Least Two Execution Units |
US8090983B2 (en) * | 2004-10-25 | 2012-01-03 | Robert Bosch Gmbh | Method and device for performing switchover operations in a computer system having at least two execution units |
US7703082B2 (en) * | 2004-12-07 | 2010-04-20 | International Business Machines Corporation | Controlling user intervention in a multi-processing computer system |
US20060123395A1 (en) * | 2004-12-07 | 2006-06-08 | International Business Machines Corporation | Controlling user intervention in a multi-processing computer system |
US7949833B1 (en) | 2005-01-13 | 2011-05-24 | Marvell International Ltd. | Transparent level 2 cache controller |
US8347034B1 (en) * | 2005-01-13 | 2013-01-01 | Marvell International Ltd. | Transparent level 2 cache that uses independent tag and valid random access memory arrays for cache access |
US8621152B1 (en) | 2005-01-13 | 2013-12-31 | Marvell International Ltd. | Transparent level 2 cache that uses independent tag and valid random access memory arrays for cache access |
KR101136610B1 (en) | 2005-04-05 | 2012-04-23 | 인텔 코오퍼레이션 | Sequencer address management |
US20090164698A1 (en) * | 2007-12-24 | 2009-06-25 | Yung-Li Ji | Nonvolatile storage device with NCQ supported and writing method for a nonvolatile storage device |
US8583854B2 (en) * | 2007-12-24 | 2013-11-12 | Skymedi Corporation | Nonvolatile storage device with NCQ supported and writing method for a nonvolatile storage device |
US20140025868A1 (en) * | 2008-10-17 | 2014-01-23 | Seagate Technology Llc | System and method for managing storage device caching |
US20130159631A1 (en) * | 2011-12-19 | 2013-06-20 | Thomas Seufert | Transactional-consistent cache for database objects |
US9152626B2 (en) * | 2011-12-19 | 2015-10-06 | Sap Se | Transactional-consistent cache for database objects |
US20140095718A1 (en) * | 2012-09-28 | 2014-04-03 | International Business Machines Corporation | Maximizing resources in a multi-application processing environment |
US20140095716A1 (en) * | 2012-09-28 | 2014-04-03 | International Business Machines Corporation | Maximizing resources in a multi-application processing environement |
US10241919B2 (en) | 2014-05-09 | 2019-03-26 | Huawei Technologies Co., Ltd. | Data caching method and computer system |
CN105094686A (en) * | 2014-05-09 | 2015-11-25 | 华为技术有限公司 | Data caching method, cache and computer system |
US20160335021A1 (en) * | 2015-05-11 | 2016-11-17 | Mentor Graphics Corporation | Memory corruption protection by tracing memory |
US9983914B2 (en) * | 2015-05-11 | 2018-05-29 | Mentor Graphics Corporation | Memory corruption protection by tracing memory |
US10678441B2 (en) | 2016-05-05 | 2020-06-09 | Micron Technology, Inc. | Non-deterministic memory protocol |
US10963164B2 (en) | 2016-05-05 | 2021-03-30 | Micron Technology, Inc. | Non-deterministic memory protocol |
US11422705B2 (en) | 2016-05-05 | 2022-08-23 | Micron Technology, Inc. | Non-deterministic memory protocol |
US11740797B2 (en) | 2016-05-05 | 2023-08-29 | Micron Technology, Inc. | Non-deterministic memory protocol |
US10534540B2 (en) | 2016-06-06 | 2020-01-14 | Micron Technology, Inc. | Memory protocol |
US11340787B2 (en) | 2016-06-06 | 2022-05-24 | Micron Technology, Inc. | Memory protocol |
US11947796B2 (en) | 2016-06-06 | 2024-04-02 | Micron Technology, Inc. | Memory protocol |
US20180292991A1 (en) * | 2017-04-11 | 2018-10-11 | Micron Technology, Inc. | Memory protocol with programmable buffer and cache size |
Also Published As
Publication number | Publication date |
---|---|
US6470422B2 (en) | 2002-10-22 |
DE19983793B4 (en) | 2013-02-07 |
GB2359163B (en) | 2003-07-30 |
AU6497199A (en) | 2000-06-26 |
US6349363B2 (en) | 2002-02-19 |
GB0111896D0 (en) | 2001-07-04 |
HK1036662A1 (en) | 2002-01-11 |
GB2359163A (en) | 2001-08-15 |
CN1179276C (en) | 2004-12-08 |
DE19983793T1 (en) | 2001-11-29 |
US20020046325A1 (en) | 2002-04-18 |
CN1334938A (en) | 2002-02-06 |
WO2000034870A1 (en) | 2000-06-15 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US6349363B2 (en) | Multi-section cache with different attributes for each section | |
US6212602B1 (en) | Cache tag caching | |
US6434669B1 (en) | Method of cache management to dynamically update information-type dependent cache policies | |
US6122709A (en) | Cache with reduced tag information storage | |
US6766419B1 (en) | Optimization of cache evictions through software hints | |
US6119205A (en) | Speculative cache line write backs to avoid hotspots | |
US6425058B1 (en) | Cache management mechanism to enable information-type dependent cache policies | |
US11030108B2 (en) | System, apparatus and method for selective enabling of locality-based instruction handling | |
US20060143390A1 (en) | Fair sharing of a cache in a multi-core/multi-threaded processor by dynamically partitioning of the cache | |
US20020116584A1 (en) | Runahead allocation protection (rap) | |
US7516276B2 (en) | Runtime register allocator | |
US6434668B1 (en) | Method of cache management to store information in particular regions of the cache according to information-type | |
US5765199A (en) | Data processor with alocate bit and method of operation | |
US9547593B2 (en) | Systems and methods for reconfiguring cache memory | |
US20200104259A1 (en) | System, method, and apparatus for snapshot prefetching to improve performance of snapshot operations | |
US11604733B1 (en) | Limiting allocation of ways in a cache based on cache maximum associativity value | |
US11256625B2 (en) | Partition identifiers for page table walk memory transactions | |
CN113874845A (en) | Multi-requestor memory access pipeline and arbiter | |
US10482017B2 (en) | Processor, method, and system for cache partitioning and control for accurate performance monitoring and optimization | |
US6360299B1 (en) | Extended cache state with prefetched stream ID information | |
US6434671B2 (en) | Software-controlled cache memory compartmentalization | |
US6421762B1 (en) | Cache allocation policy based on speculative request history | |
US8266379B2 (en) | Multithreaded processor with multiple caches | |
US6510494B1 (en) | Time based mechanism for cached speculative data deallocation | |
WO1997034229A9 (en) | Segment descriptor cache for a processor |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
AS | Assignment |
Owner name: INTEL CORPORATION, CALIFORNIA Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:CAI, ZHONG-NING;NAKANISHI, TOSAKU;REEL/FRAME:009637/0339;SIGNING DATES FROM 19981124 TO 19981130 |
|
STCF | Information on status: patent grant |
Free format text: PATENTED CASE |
|
FPAY | Fee payment |
Year of fee payment: 4 |
|
FPAY | Fee payment |
Year of fee payment: 8 |
|
FEPP | Fee payment procedure |
Free format text: PAYOR NUMBER ASSIGNED (ORIGINAL EVENT CODE: ASPN); ENTITY STATUS OF PATENT OWNER: LARGE ENTITY |
|
FPAY | Fee payment |
Year of fee payment: 12 |