WO2022261223A1 - Apparatus, system, and method for configuring a configurable combined private and shared cache - Google Patents
Apparatus, system, and method for configuring a configurable combined private and shared cache Download PDFInfo
- Publication number
- WO2022261223A1 WO2022261223A1 PCT/US2022/032694 US2022032694W WO2022261223A1 WO 2022261223 A1 WO2022261223 A1 WO 2022261223A1 US 2022032694 W US2022032694 W US 2022032694W WO 2022261223 A1 WO2022261223 A1 WO 2022261223A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- private
- cache
- shared
- processor
- shared cache
- Prior art date
Links
- 238000000034 method Methods 0.000 title claims abstract description 21
- 238000004891 communication Methods 0.000 claims description 6
- 230000008859 change Effects 0.000 claims description 4
- 230000001413 cellular effect Effects 0.000 claims description 2
- 230000036541 health Effects 0.000 claims description 2
- 230000000977 initiatory effect Effects 0.000 claims description 2
- 238000012545 processing Methods 0.000 description 28
- 238000010586 diagram Methods 0.000 description 8
- 230000004044 response Effects 0.000 description 8
- 238000013461 design Methods 0.000 description 6
- 230000006870 function Effects 0.000 description 5
- 238000003860 storage Methods 0.000 description 5
- 238000013459 approach Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 239000002245 particle Substances 0.000 description 2
- 230000003068 static effect Effects 0.000 description 2
- 238000003491 array Methods 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 230000001427 coherent effect Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 239000004744 fabric Substances 0.000 description 1
- 239000004973 liquid crystal related substance Substances 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
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
- G06F12/0806—Multiuser, multiprocessor or multiprocessing cache systems
- G06F12/084—Multiuser, multiprocessor or multiprocessing cache systems with a shared cache
-
- 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
- G06F2212/00—Indexing scheme relating to accessing, addressing or allocation within memory systems or architectures
- G06F2212/60—Details of cache memory
- G06F2212/601—Reconfiguration of cache memory
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2212/00—Indexing scheme relating to accessing, addressing or allocation within memory systems or architectures
- G06F2212/60—Details of cache memory
- G06F2212/6042—Allocation of cache space to multiple users or processors
Definitions
- the technology of the disclosure relates generally to configuring cache resources in a microprocessor, and specifically to configuring combined private and shared cache levels in a microprocessor.
- microprocessors may conventionally include one or more levels of cache memory resources. These cache memory resources may be arranged in a hierarchical manner - for example, a microprocessor may have level 0 (L0), level 1 (LI), level 2 (L2), and level 3(L3) caches.
- L0 cache may be the relative smallest and lowest latency, with the other caches increasing in size and latency up through the L3 cache, which may be the largest but with the longest latency compared to the other caches.
- one or more of the levels of cache hierarchy may have split instruction and data caches (e.g., the L0 cache level may comprise split L0 instruction and L0 data caches), whereas other levels of the cache hierarchy may contain both instructions and data.
- Some levels of the cache hierarchy may be “private” to the microprocessor or, in the case of a multi-core microprocessor, may be private to one or more individual core(s) (meaning that such private caches are only visible and accessible to the associated microprocessor or individual core(s)).
- Other levels of the cache hierarchy despite being physically located with a particular microprocessor, may be shared across and usable by one or more other microprocessors in a system.
- a conventional approach to cache design is to provide each core of the microprocessor with a private L2 cache of a fixed size, and a shared L3 of a fixed size which is shared among the cores of the microprocessor (and potentially across a system-on-chip interconnect).
- this approach can provide some benefit to many types of workloads, it does not allow for optimization to fit the characteristics of a specific workload.
- the static cache allocation described above may lead to lower overall performance of the workload (because the workload cannot use the existing cache hierarchy efficiently), or lower overall utilization of the available computing resources (if fewer instances of the workload are run per microprocessor to alleviate the performance problems as a result of the static cache hierarchy).
- aspects disclosed in the detailed description include configuring a configurable combined private and shared cache in a processor.
- the configurable combined private and shared cache can function as a physically combined but logically separated private and shared cache for providing cache memory access to a processor and/or its processing cores.
- the configurable combined private and shared cache includes a plurality of cache ways. Because the private and shared portions of the configurable combined private and shared cache are included in the same physical cache structure, it is possible to change the relative amounts of the combined private and shared cache that are devoted to private and shared portions respectively.
- a configuration is programmed or set in the configurable combined private and shared cache which allocates a first portion of the cache ways to the shared L3 portion, and a second portion of the cache ways to the private portion.
- the configuration of the combined private and shared cache can be dynamic and may be changed during operation of the processor, for example as part of a firmware configuration, by boot-time configuration, during system resets, during operation of the processor when the contents of all caches are to be flushed or invalidated, or at other times and in other manners that will occur to those having skill in the art.
- the configuration may be changed over time, as the processor-based system may from time to time have different applications or virtual machines allocated to ran on the processor.
- a processor-based system comprising a configurable combined private and shared cache, the configurable combined private and shared cache configured to include a shared portion and a private portion, and to be responsive to a dynamic cache configuration determining a size of each of the shared portion and the private portion.
- a processor-based system comprises a means for storing information, the means include a shared portion and a private portion, and responsive to a means for configuring the means for storing information determining a size of each of the shared portion and the private portion
- a method of configuring caches in a processor-based system comprises dynamically configuring a combined private and shared cache structure to select a private cache portion and a shared cache portion.
- FIG. 1 is a block diagram of an exemplary system for configuring a combined private and share cache in a processor-based system
- Figure 2 is a block diagram of another exemplary system including combined private and shared caches in a processor-based system
- Figure 3 is a block diagram of a method for configuring a combined private and shared cache
- Figure 4 illustrates an exemplary configuration register (CR) that may be provided in some aspects to enable configuration of combined private and shared caches
- Figure 5 illustrates another exemplary CR that may be provided for configuration of combined private and shared caches according to some aspects
- Figures 6A and 6B illustrates exemplary communications flows and operations performed by processor cores and home nodes for sending and responding to cache requests according to some aspects
- Figure 7 is a block diagram of an exemplary processor-based system including a processor having a combined private and shared cache.
- aspects disclosed in the detailed description include configuring a configurable combined private and shared cache in a processor.
- the configurable combined private and shared cache can function as a physically combined but logically separated private and shared cache for providing cache memory access to a processor and/or its processing cores.
- the configurable combined private and shared cache includes a plurality of cache ways. Because the private and shared portions of the configurable combined private and shared cache are included in the same physical cache structure, it is possible to change the relative amounts of the combined private and shared cache that are devoted to private and shared portions respectively.
- a configuration is programmed or set in the configurable combined private and shared cache which allocates a first portion of the cache ways to the shared L3 portion, and a second portion of the cache ways to the private portion.
- the configuration of the combined private and shared cache can be dynamic and may be changed during operation of the processor, for example as part of a firmware configuration, by boot-time configuration, during system resets, during operation of the processor when the contents of all caches are to be flushed or invalidated, or at other times and in other manners that will occur to those having skill in the art.
- the configuration may be changed over time, as the processor-based system may from time to time have different applications or virtual machines allocated to ran on the processor.
- FIG. 1 is a block diagram of a processor-based system 100 including a processor 105 (e.g., a microprocessor) including a configurable combined private and shared cache, as will be discussed in greater detail herein.
- the processor 105 includes a first processing core 121 and a second processing core 131.
- the first processing core 121 further includes execution circuits 122, which may also be referred to as core logic circuitry 122, which are coupled to low level cache(s) 123 and a configurable combined private and shared cache 124, and which may receive instructions and data on which to perform operations from the low level cache(s) 123 and the combined private and shared cache 124.
- the low level cache(s) 123 may include L0 and LI caches (whether split data and instruction caches, or combined), and the configurable combined private and shared cache 124 may include a private L2 and a shared L3 cache, in an aspect.
- the second processing core 131 further includes execution circuits 132, which may also be referred to as core logic circuitry 132, which are coupled to low level cache(s) 133 and a configurable combined private and shared cache 134, and which may receive instructions and data on which to perform operations from the low level cache(s) 133 and the combined private and shared cache 134.
- the low level cache(s) 133 may include L0 and LI caches (whether split data and instruction caches, or combined), and the configurable combined private and shared cache 134 may include a private L2 and a shared L3 cache, in an aspect.
- the first processing core 121 and the second processing core 131 may communicate over an interconnect 140.
- the configurable combined private and shared cache 124 functions as a physically combined but logically separated private L2 and shared L3 cache, illustrated as a shared L3 portion 124a and private L2 portion 124b, and includes a plurality of cache ways 125-1 through 125-12.
- the private L2 and shared L3 portions are included in the same physical cache structure, it is possible to change the relative amounts of the combined private and shared cache 124 that are devoted to private L2 and shared L3 portions respectively.
- a configuration is programmed or set in the configurable combined private and shared cache 124 which allocates a first portion of the cache ways 125-1 to 125-12 to the shared L3 portion 124a, and a second portion of the cache ways 125-1 to 125-12 to the private L2 portion 124b.
- the configuration of the combined private and shared cache 124 is dynamic and may be changed during operation of the processor 105, for example as part of a firmware configuration, by boot-time configuration, during system resets, during operation of the processor 105 when the contents of all caches are to be flushed or invalidated, or at other times and in other manners that will occur to those having skill in the art.
- the configuration may be changed over time, as the processor-based system 100 may from time to time have different applications or virtual machines allocated to run on the processor 105.
- the shared L3 portion 124a may include cache ways 125-1 to 125-6, while the private L2 portion 124b may include cache ways 125-7 to 125-12.
- the configuration may be changed such that the shared L3 portion 124 may include cache ways 125-1 to 125-8, while the private L2 portion 124b may include cache ways 125-9 to 125-12.
- the configurable combined private and shared cache 124 may allow complete flexibility regarding the sizes of the shared L3 portion 124a and the private L2 portion 124b (i.e., each of the portions may be set to a size anywhere from zero ways to all ways of the configurable combined private and shared cache 124), whereas in other aspects, lower and/or upper limits on the size of either or both the shared L3 portion 124a and the private L2 portion 124b may be established (e.g., the private L2 portion 124b may not be smaller than two ways, and/or the shared L3 portion 124a may not be smaller than four ways).
- the configurable combined private and shared cache 134 may function similarly to the configurable combined private and shared cache 124, and as such may also include a plurality of cache ways 125-1 through 125-12.
- the configurable combined private and shared cache 134 may share a configuration with the configurable combined private and shared cache 124, or may be configured differently, depending on any requirements of an associated system architecture, design considerations, or other design choices as will be apparent to those having skill in the art.
- FIG. 2 is a block diagram of a processor-based system 200 including combined private and shared caches.
- the processor-based system 200 includes a first processing cluster 211, which is coupled to a first core cluster interface 217 and a first router 218. A first home node 219 is coupled to the first router 218.
- the processor- based system 200 further includes a second processing cluster 221, which is coupled to a second core cluster interface 227 and a second router 228.
- a second home node 229 is coupled to the second router 228.
- the processor-based system 200 further includes a third processing cluster 231, which is coupled to a third core cluster interface 237 and a third router 238.
- a third home node 239 is coupled to the third router 238.
- the home nodes 219, 229, and 239 may manage coherency for at least a set of memory addresses of the processor-based system 200 for their associated processing clusters, which may include snooping and invalidating caches, allocating and managing the shared caches, and performing transactions with a system memory (e.g., DDR DRAM in one aspect).
- the routers 218, 228, and 238 may be coupled together as part of a system interconnect in order to allow the processing clusters 211, 221, and 231 to communicate and transfer instructions and data between them and to access system memory.
- the first processing cluster 211 includes four processing cores 21 la-21 Id. Each of the processing cores 21 la-21 Id includes a configurable combined private and shared cache which has been configured to include ten (10) ways of private cache and 2 ways of shared cache.
- the second processing cluster 221 includes four processing cores 221a-221d. Each of the processing cores 221a-221d includes a configurable combined private and shared cache which has been configured to include 6 ways of private cache and 6 ways of shared cache.
- the third processing cluster 231 includes four processing cores 231a-231d, but two of the processing cores (231b and 231d) are disabled (e.g., as a result of manufacturing defects, by deliberate fusing, or other configuration methods), although their caches are still enabled and available.
- FIG. 3 is a block diagram of a method 300 for configuring a combined private and shared cache according to one aspect.
- the method 300 begins in block 310 by dynamically configuring a combined private and shared cache structure to select a private cache portion and a shared cache portion.
- this configuration may be included in a firmware configuration, performed by boot- time configuration, during system resets, or during operation of the processor 105 when the contents of all caches are to be flushed or invalidated.
- the combined private and shared cache structure comprises a plurality of ways
- the operation of block 310 for dynamically configuring the combined private and shared cache structure to select the private cache portion and the shared cache portion may comprise allocating the shared cache portion and the private cache portion to include at least a respective shared portion and a respective private portion of the plurality of ways.
- operations for allocating the shared cache portion and the private cache portion to include at least a respective shared portion and a respective private portion of the plurality of ways may comprise allocating either of the shared cache portion and the private cache portion to include all of the plurality of ways.
- Some aspects may provide that the operations of block 310 for dynamically configuring the combined private and shared cache structure to select the private cache portion and the shared cache portion may comprise storing a value in a configuration register (CR) that indicates a count of the plurality of ways assigned as the shared cache portion.
- CR configuration register
- the method 300 optionally continues at block 320 by changing the configuration of the combined private and shared cache structure to select a second private cache portion different than the private cache portion, and a second shared cache portion different that the shared cache portion.
- the shared L3 portion 124a may include cache ways 125-1 to 125-6, while the private L2 portion 124b may include cache ways 125-7 to 125-12.
- the configuration may be changed such that the shared L3 portion 124 may include cache ways 125-1 to 125-8, while the private L2 portion 124b may include cache ways 125-9 to 125-12.
- Figures 4 and 5 illustrate registers that may be provided in some aspects to enable configuration of combined private and shared caches such as the configurable combined private and shared caches 124 and 134 of Figure 1.
- a Core_CPU_Cache_Config register 400 is a 32-bit CR that may be implemented in each CPU core of a processor-based device, such as the processing cores 211a-211d, 221a-221d, and 231a-231d of the processor-based device 200 of Figure 2.
- the Core_CPU_Cache_Config register 400 comprises a CPU_Cache_Ways field 402 and a CPU_Cache_L3_Ways field 404, each of which are five (5) bits in size, with the remainder of the CPU_Cache_Config register 400 comprising a reserved space 406.
- the CPU_Cache_Ways field 402 stores a read-only value that indicates to firmware and/or software a total number of ways in each set of a corresponding CPU cache.
- the CPU_Cache_L3_Ways field 404 stores a read/write value that enables the firmware and/or software to configure how many of the ways are assigned as shared L3 cache. The remaining ways in the CPU cache are available to the CPU core as private L2 cache.
- the Core_CPU_Cache_Config register 400 applies to all sets in the CPU cache.
- FIG. 5 shows a Home_Node_CPU_[O..N]_Cache_Config register 500, which is a 32-bit CR implemented in each home node of a processor-based device (e.g., the home nodes 219, 229, and 239 of the processor-based device 200 of Figure 2), with one copy of the Home_Node_CPU_[O..N]_Cache_Config register 500 implemented for each CPU core (e.g., the processing cores 21 la-21 Id, 221a-221d, and 23 la-23 Id of Figure 2) of the processor-based device.
- a processor-based device e.g., the home nodes 219, 229, and 239 of the processor-based device 200 of Figure 2
- the Home_Node_CPU_[O..N]_Cache_Config register 500 implemented for each CPU core (e.g., the processing cores 21 la-21 Id, 221a-221d, and 23 la-23 Id of Figure 2) of
- the Home_Node_CPU_[O..N]_Cache_Config register 500 comprises a CPU_Cache_L3_Ways field 502 that is five (5) bits in size, and that stores a read/write value that allows firmware and/or software to configure how many of the ways of each CPU core are assigned as shared L3 cache. This value is used by the corresponding home node to build a weighting table (not shown), in which CPU cores with more assigned L3 ways have a higher weighting than CPU cores with fewer assigned L3 ways.
- the CPU_Cache_L3_Ways field 502 for each CPU core must be configured with the same value as the CPU_Cache_L3_Ways field 404 in the Core_CPU_Cache_Config register 400 in the corresponding CPU core.
- Figure 6A illustrates a transaction, based on the ARM Coherent Hub Interface (CHI) coherency protocol, in which a read to an address A misses on a first processor core’s own CPU cache but hits in a shared L3 portion of a second processor core, resulting in allocation in the private L2 cache of the second processor core and eviction of an existing entry for an address B.
- Figure 6B illustrates a transaction involving the writeback of address B (including dirty data), which is then allocated in the shared L3 cache of the second processor core.
- CHI ARM Coherent Hub Interface
- Figures 6A and 6B include a processor core designated as Core-Al 600, a corresponding home node designated as Home Node-Al 602, a processor core designated as Core-A2 604, and a corresponding home node designated as Home Node-A2 606.
- Each of the Core-Al 600, the Home Node-Al 602, the Core-A2 604 and the Home Node-A2 606 are represented by vertical lines, with operations performed by each element represented by blocks on the corresponding vertical line and communications between elements represented by arrows between the corresponding vertical lines.
- the Core-Al 600 needs to read address A with unique ownership (e.g., so that it can execute a store instruction to address A), so it performs a ReadUnique request (i.e., a request to read a memory location and obtain exclusive ownership of the corresponding cache line) to address A, which is routed to the Home Node-Al 602, as indicated by arrow 612.
- the Home Node-Al 602 determines that the Core-A2604 is holding a unique copy of the contents of address A, and thus the Home Node-Al 602 sends a forwarding snoop (i.e., an invalidating snoop including a request to forward data to the requestor) to the Core-A2 604, as indicated by arrow 614.
- a forwarding snoop i.e., an invalidating snoop including a request to forward data to the requestor
- the Core-A2 604 sends a SnpResp_I_Fwded_UC[A] response to the Home Node- A 1 602 to inform it that it has forwarded the cache line to the requester in the UC state and invalidated its copy, as indicated by arrow 616.
- the Core-A2 604 also invalidates its copy of the contents of address A (block 618).
- the Core-A2 604 then sends a CompData_UC message (i.e., a combined completion and data message along with the UC state) to forward the cache line of address A to the Core-Al 600, as indicated by arrow 620.
- CompData_UC message i.e., a combined completion and data message along with the UC state
- the Core-Al 600 allocates space for and stores the contents of address A in its L2 cache, where it holds the contents of address A in a UC state (block 622).
- the Core-Al-600 then sends a final CompAck message (i.e., a completion acknowledgement message) response to the Home Node-Al 602 to acknowledge that it has received the cache line and the transaction is complete, as indicated by arrow 624.
- a final CompAck message i.e., a completion acknowledgement message
- the Core-Al 600 evicts the contents of address B from its L2 cache and needs to write it back to memory, so the Core-Al 600 sends a WriteBackFullfB] request (i.e., a request to write back a full cache line of dirty data from the cache) to the Home Node-A2 606, as indicated by arrow 626.
- the Home Node-A2 606 selects the Core-A2 604 to allocate space in its L3 cache for the evicted contents of address B (block 628).
- the Home Node-S2 606 then sends a SnpStashShared request (i.e., a snoop message with a stash request that the cache line be read and allocated in the cache) to the Core-A2 604, as indicated by arrow 630.
- the Core-A2 604 then sends a SnpResp_I_Read[B] response (i.e., a snoop response indicating that the cache line is not currently valid, and including a red request to obtain the cache line) to the Home Node- A2606 in response to the stashing snoop request, as indicated by arrow 632.
- the Home Node-A2 606 next sends a CompDBIDResp[B] response (i.e., a combined completion and data buffer ID response) to informing the Core-Al 600 that it can proceed with sending the writeback data to the Core-A2604 using the data buffer ID provided), as indicated by arrow 634.
- the Core-Al 600 invalidates its copy of the contents of address B in its L2 cache (block 636).
- the Core-Al 600 then sends a CBWrData_UD_PD message (i.e., a message to copy back write data for the writeback, in which the cache line was in the UD state when it was sent, and the UD state is being passed to the target) to the CoreOA2 604, as indicated by arrow 638.
- the Core-A2 604 allocates space for and stores the contents of address B in its L3 cache (i.e., based on a dynamic configuration defining its L3 cache) (block 640).
- the Core-A2604 sends a final CompAck message (i.e., a completion acknowledgement message) response to the Home Node-A2 606 to inform it that it has received the cache line, and the transaction is complete, as indicated by arrow 642.
- the exemplary processor including a configurable combined private and shared cache structure may be provided in or integrated into any processor-based device.
- Examples include a server, a computer, a portable computer, a desktop computer, a mobile computing device, a set top box, an entertainment unit, a navigation device, a communications device, a fixed location data unit, a mobile location data unit, a global positioning system (GPS) device, a mobile phone, a cellular phone, a smart phone, a session initiation protocol (SIP) phone, a tablet, a phablet, a wearable computing device (e.g., a smart watch, a health or fitness tracker, eyewear, etc.), a personal digital assistant (PDA), a monitor, a computer monitor, a television, a tuner, a radio, a satellite radio, a music player, a digital music player, a portable music player, a digital video player, a video player
- PDA personal digital assistant
- Figure 7 illustrates an example of a processor-based system 700 that includes a configurable combined private and shared cache structure as illustrated and described with respect to Figures 1-6.
- the processor-based system 700 includes a processor 701 having one or more central processing units (CPUs) 705, each including one or more processor cores, and which may correspond to the processor 105 of Figure 1, and as such may include a configurable combined private and shared cache structure as illustrated and described with respect to Figures 1-6.
- the CPU(s) 705 may be a master device.
- the CPU(s) 705 is coupled to a system bus 710 and can intercouple master and slave devices included in the processor-based system 700.
- the CPU(s) 705 communicates with these other devices by exchanging address, control, and data information over the system bus 710.
- the CPU(s) 705 can communicate bus transaction requests to a memory controller 751 as an example of a slave device.
- a memory controller 751 as an example of a slave device.
- multiple system buses 710 could be provided, wherein each system bus 710 constitutes a different fabric.
- Other master and slave devices can be connected to the system bus 710. As illustrated in Figure 7, these devices can include a memory system 750, one or more input devices 420, one or more output devices 430, one or more network interface devices 740, and one or more display controllers 760, as examples.
- the input device(s) 720 can include any type of input device, including, but not limited to, input keys, switches, voice processors, etc.
- the output device(s) 730 can include any type of output device, including, but not limited to, audio, video, other visual indicators, etc.
- the network interface device(s) 740 can be any devices configured to allow exchange of data to and from a network 745.
- the network 745 can be any type of network, including, but not limited to, a wired or wireless network, a private or public network, a local area network (LAN), a wireless local area network (WLAN), a wide area network (WAN), a BLUETOOTHTM network, and the Internet.
- the network interface device(s) 740 can be configured to support any type of communications protocol desired.
- the memory system 750 can include the memory controller 751 coupled to one or more memory arrays 752.
- the CPU(s) 705 may also be configured to access the display controller(s) 760 over the system bus 710 to control information sent to one or more displays 762.
- the display controller(s) 760 sends information to the display(s) 762 to be displayed via one or more video processors 761, which process the information to be displayed into a format suitable for the display(s) 762.
- the display(s) 762 can include any type of display, including, but not limited to, a cathode ray tube (CRT), a liquid crystal display (LCD), a plasma display, a light emitting diode (LED) display, etc.
- CTR cathode ray tube
- LCD liquid crystal display
- LED light emitting diode
- DSP Digital Signal Processor
- ASIC Application Specific Integrated Circuit
- FPGA Field Programmable Gate Array
- a processor may be a microprocessor, but in the alternative, the processor may be any conventional processor, controller, microcontroller, or state machine.
- a processor may also be implemented as a combination of computing devices (e.g., a combination of a DSP and a microprocessor, a plurality of microprocessors, one or more microprocessors in conjunction with a DSP core, or any other such configuration).
- RAM Random Access Memory
- ROM Read Only Memory
- EPROM Electrically Programmable ROM
- EEPROM Electrically Erasable Programmable ROM
- registers a hard disk, a removable disk, a CD-ROM, or any other form of computer readable medium known in the art.
- An exemplary storage medium is coupled to the processor such that the processor can read information from, and write information to, the storage medium.
- the storage medium may be integral to the processor.
- the processor and the storage medium may reside in an ASIC.
- the ASIC may reside in a remote station.
- the processor and the storage medium may reside as discrete components in a remote station, base station, or server.
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Memory System Of A Hierarchy Structure (AREA)
Abstract
Description
Claims
Priority Applications (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
EP22740693.1A EP4352621A1 (en) | 2021-06-09 | 2022-06-08 | Apparatus, system, and method for configuring a configurable combined private and shared cache |
CN202280055467.9A CN117795491A (en) | 2021-06-09 | 2022-06-08 | Apparatus, system, and method for configuring a configurable private and shared combined cache |
Applications Claiming Priority (4)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US202163208722P | 2021-06-09 | 2021-06-09 | |
US63/208,722 | 2021-06-09 | ||
US17/834,661 US11880306B2 (en) | 2021-06-09 | 2022-06-07 | Apparatus, system, and method for configuring a configurable combined private and shared cache |
US17/834,661 | 2022-06-07 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2022261223A1 true WO2022261223A1 (en) | 2022-12-15 |
Family
ID=82482932
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/US2022/032694 WO2022261223A1 (en) | 2021-06-09 | 2022-06-08 | Apparatus, system, and method for configuring a configurable combined private and shared cache |
Country Status (1)
Country | Link |
---|---|
WO (1) | WO2022261223A1 (en) |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5875464A (en) * | 1991-12-10 | 1999-02-23 | International Business Machines Corporation | Computer system with private and shared partitions in cache |
US20070143546A1 (en) * | 2005-12-21 | 2007-06-21 | Intel Corporation | Partitioned shared cache |
US20120198172A1 (en) * | 2009-08-25 | 2012-08-02 | International Business Machines Corporation | Cache Partitioning in Virtualized Environments |
WO2021066844A1 (en) * | 2019-10-04 | 2021-04-08 | Visa International Service Association | Techniques for multi-tiered data storage in multi-tenant caching systems |
-
2022
- 2022-06-08 WO PCT/US2022/032694 patent/WO2022261223A1/en active Application Filing
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5875464A (en) * | 1991-12-10 | 1999-02-23 | International Business Machines Corporation | Computer system with private and shared partitions in cache |
US20070143546A1 (en) * | 2005-12-21 | 2007-06-21 | Intel Corporation | Partitioned shared cache |
US20120198172A1 (en) * | 2009-08-25 | 2012-08-02 | International Business Machines Corporation | Cache Partitioning in Virtualized Environments |
WO2021066844A1 (en) * | 2019-10-04 | 2021-04-08 | Visa International Service Association | Techniques for multi-tiered data storage in multi-tenant caching systems |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
JP7116047B2 (en) | Memory controller and method for flexible management of heterogeneous memory systems in processor-based systems | |
US7529894B2 (en) | Use of FBDIMM channel as memory channel and coherence channel | |
JP2009037615A (en) | Inclusive shared cache among a plurality of core-cache clusters | |
JP6262407B1 (en) | Providing shared cache memory allocation control in shared cache memory systems | |
US7353340B2 (en) | Multiple independent coherence planes for maintaining coherency | |
US7398360B2 (en) | Multi-socket symmetric multiprocessing (SMP) system for chip multi-threaded (CMT) processors | |
CN115408307A (en) | Cache for storing coherent data and non-coherent data | |
US12007896B2 (en) | Apparatuses, systems, and methods for configuring combined private and shared cache levels in a processor-based system | |
US11507517B2 (en) | Scalable region-based directory | |
EP3475832A1 (en) | Self-aware, peer-to-peer cache transfers between local, shared cache memories in a multi-processor system | |
WO2019055599A1 (en) | Converting a stale cache memory unique request to a read unique snoop response in a multiple (multi-) central processing unit (cpu) processor to reduce latency associated with reissuing the stale unique request | |
US11880306B2 (en) | Apparatus, system, and method for configuring a configurable combined private and shared cache | |
WO2022261223A1 (en) | Apparatus, system, and method for configuring a configurable combined private and shared cache | |
US11947454B2 (en) | Apparatuses, systems, and methods for controlling cache allocations in a configurable combined private and shared cache in a processor-based system | |
WO2017146882A1 (en) | Providing scalable dynamic random access memory (dram) cache management using dram cache indicator caches | |
WO2022261226A1 (en) | Apparatuses, systems, and methods for configuring combined private and shared cache levels in a processor-based system | |
CN117795490A (en) | Apparatus, system, and method for configuring combined private and shared cache hierarchies in a processor-based system | |
EP4352622A1 (en) | Apparatus and method for controlling allocations in a shared cache of a numa system | |
US20240095173A1 (en) | Providing fairness-based allocation of caches in processor-based devices | |
CN117813593A (en) | Apparatus and method for controlling allocation in shared cache of NUMA system | |
JP6396625B1 (en) | Maintaining cache coherency using conditional intervention between multiple master devices | |
WO2024129347A1 (en) | Processor-based system for allocating cache lines to a higher-level cache memory | |
WO2024086437A1 (en) | Page rinsing scheme to keep a directory page in an exclusive state in.a single complex | |
US20190065372A1 (en) | Providing private cache allocation for power-collapsed processor cores in processor-based systems | |
WO2007022375A2 (en) | Multiple independent coherence planes for maintaining coherency |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 22740693 Country of ref document: EP Kind code of ref document: A1 |
|
DPE1 | Request for preliminary examination filed after expiration of 19th month from priority date (pct application filed from 20040101) | ||
WWE | Wipo information: entry into national phase |
Ref document number: 2022740693 Country of ref document: EP |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
ENP | Entry into the national phase |
Ref document number: 2022740693 Country of ref document: EP Effective date: 20240109 |
|
WWE | Wipo information: entry into national phase |
Ref document number: 202280055467.9 Country of ref document: CN |