Advanced Micro Devices Corporation

Matthew R Poremba

Gabriel H Loh

Jagadish B Kotra

John Kalamatianos

Yasuko Eckert

Sooraj Puthoor

Steven E Raasch

Alexandru Dutu

Marko Scrbak

Jieming Yin

Adithya Yalavarti

Doug Hunt

Ersin Cukurtas

An approach provides indirect addressing support for PIM. Indirect PIM commands include address translation information that allows memory modules to perform indirect addressing. Processing logic in a memory module processes an indirect PIM command and retrieves, from a first memory location, a virtual address of a second memory location. The processing logic calculates a corresponding physical address for the virtual address using the address translation information included with the indirect PIM command and retrieves, from the second memory location, a virtual address of a third memory location. This process is repeated any number of times until one or more indirection stop criteria are satisfied. The indirection stop criteria stop the process when work has been completed normally or to prevent errors. Implementations include the processing logic in the memory module working in cooperation with a memory controller to perform indirect addressing.

Approach for providing indirect addressing in memory modules

Sparse matrix operations using processing-in-memory is described. In accordance with the described techniques, a processing-in-memory component of a memory module receives a request for a vector element stored at a first location in memory of the memory module. The processing-in-memory component identifies an index value for a non-zero element in a sparse matrix using a representation of the sparse matrix stored at a second location in the memory. The processing-in-memory component then outputs a result that includes the vector element by retrieving the vector element from the first location in memory using the index value.

Sparse matrix operations using processing-in-memory

In accordance with described techniques for processing-in-memory (PIM) search stop control, a computing system or computing device includes a memory system that includes a stop condition check component, which receives an instruction that includes a programmed check value. The stop condition check component compares the programmed check value to outputs of a PIM component, and the stop condition check component initiates a stop instruction to stop the PIM component from processing subsequent computations based on an output of the PIM component matching the programmed check value.

PIM search stop control

Devices and methods for cache prefetching are provided. A device is provided which comprises memory and a processor. The memory comprises a DRAM cache, a cache dedicated to the processor and one or more intermediate caches between the dedicated cache and the DRAM cache. The processor is configured to issue prefetch requests to prefetch data, issue data access requests to fetch the data and when one or more previously issued prefetch requests are determined to be inaccurate, issue a prefetch request to prefetch a tag, corresponding to the memory address of requested data in the DRAM cache. A tag look-up is performed at the DRAM cache without performing tag look-ups at the dedicated cache or the intermediate caches. The tag is prefetched from the DRAM cache without prefetching the requested data.

Method and apparatus for a dram cache tag prefetcher

An electronic device is described that handles control transfer instructions (CTIs) when executing instructions in program code. The electronic device has a processor that includes a branch prediction functional block and a sequential fetch logic functional block. The sequential fetch logic functional block determines, based on a record associated with a CTI, that a specified number of fetch groups of instructions that were previously determined to include no CTIs are to be fetched for execution in sequence following the CTI. When each of the specified number of fetch groups is fetched and prepared for execution, the sequential fetch logic prevents corresponding accesses of the branch prediction functional block for acquiring branch prediction information for instructions in that fetch group.

Controlling accesses to a branch prediction unit for sequences of fetch groups

A processing system selectively compresses cache lines at a cache or at a memory or encrypts cache lines at the memory based on evictions of entries mapping virtual-to-physical address translations from a translation lookaside buffer (TLB). Upon eviction of a TLB entry, the processing system identifies cache lines corresponding to the physical addresses of the evicted TLB entry and selectively compresses the cache lines to increase the effective storage capacity of the processing system or encrypts the cache lines to protect against vulnerabilities.

Data compression and encryption based on translation lookaside buffer evictions

An electronic device handles memory access requests for data in a memory. The electronic device includes a memory controller for the memory, a last-level cache memory, a request generator, and a predictor. The predictor determines a likelihood that a cache memory access request for data at a given address will hit in the last-level cache memory. Based on the likelihood, the predictor determines: whether a memory access request is to be sent by the request generator to the memory controller for the data in parallel with the cache memory access request being resolved in the last-level cache memory, and, when the memory access request is to be sent, a type of memory access request that is to be sent. When the memory access request is to be sent, the predictor causes the request generator to send a memory request of the type to the memory controller.

Using predictions of outcomes of cache memory access requests for controlling whether a request generator sends memory access requests to a memory in parallel with cache memory access requests

A type of conditional probability fetcher prefetches data, such as for a cache, from another memory by maintaining information relating to memory elements in a group of memory elements fetched from the second memory. The information may be an aggregate number of memory elements that have been fetched for different memory segments in the group. The information is maintained responsive to fetching one or more memory elements from a segment of memory elements in the group of memory elements. Prefetching one or more remaining memory elements in a particular segment of memory elements from the second memory into the first memory occurs when the information relating to the memory elements in the group of memory elements indicates that a prefetching condition has been satisfied.

Conditional prefetching

Data caching methods and systems are provided. A method is provided for a hybrid cache system that dynamically changes modes of one or more cache rows of a cache between an un-split mode having a first tag field and a first data field to a split mode having a second tag field, a second data field being smaller than the first data field and a mapped page field to improve the cache access efficiency of a workflow being executed in a processor. A hybrid cache system is provided in which the cache is configured to operate one or more cache rows in an un-split mode or in a split mode. The system is configured to dynamically change modes of the cache rows from the un-split mode to the split mode to improve the cache access efficiency of a workflow being executed by the processor.

Hybrid cache

Data caching methods and systems are provided. The data cache method loads data into an installation cache and a cache (simultaneously or serially) and returns data from the installation cache when the data has not completely loaded into the cache. The data cache system includes a processor, a memory coupled to the processor, a cache coupled to the processor and the memory and an installation cache coupled to the processor and the memory. The system is configured to load data from the memory into the installation cache and the cache (simultaneously or serially) and return data from the installation cache to the processor when the data has not completely loaded into the cache.

Growing community of inventors

Shavertown, PA, United States of America

Matthew R Poremba