Intel Corporation

Ruchira Sasanka

Jayaram Bobba

David J Sager

Jeffrey J Cook

Suresh Srinivas

Rajat Agarwal

Alexander Gendler

Udi Sherel

Abhinav Das

Youfeng Wu

Koichi Yamada

Ron Gabor

Joseph Nuzman

Shlomo Raikin

Tin-Fook Ngai

Arvind Krishnaswamy

Ho-Seop Kim

Omar M Shaikh

Leeor Peled

Howard Chen

Jason A Domer

Elmoustapha Ould-Ahmed-Vall

Jason M Agron

Michael R Greenfield

Karthik Raman

Konstantinos Krommydas

Jeffery J Cook

An apparatus and method to reduce bandwidth and latency associated with probabilistic caches. For example, one embodiment of a processor comprises: a plurality of cores to execute instructions and process data, one or more of the cores to generate a request for a first cache line; a cache controller comprising cache lookup logic to determine a first way of a cache in which to search for the first cache line based on a first set of tag bits comprising one or more bits associated with the first cache line; the cache lookup logic to compare a second set of tag bits of the first cache line with a third set of tag bits of an existing cache line stored in the first way, wherein if the second set of tag bits and the third set of tag bits to not match, then the cache lookup logic to determine that the first cache line is not in the first way and to compare a fourth set of tag bits of the first cache line with a fifth set of tag bits of the existing cache line, wherein responsive to a match between the fourth set of tag bits and the fifth set of tag bits, the cache lookup logic to determine that the first cache line is stored in a second way and to responsively read the first cache line from the second way.

Apparatus and method to reduce bandwidth and latency overheads of probabilistic caches

A processor includes a core to execute a transaction with a memory via cache; and cache controller having an index mapper circuit to: identify a physical memory address associated with the transaction and having a plurality of bits; determine, based on the plurality of bits, a first set of bits encoding a tag value, a second set of bits encoding a page index value, and a third set of bits encoding a line index value; determine a mapping function corresponding to the tag value; determine, using the mapping function, a bit-placement order; combine, based on the order, second and third set of bits to form an index; generate, using the index, a mapping from the address to a cache line index value identifying a cache line in the cache; and wherein the cache controller is further to access, using the mapping and in response to the transaction, the cache line.

Reducing conflicts in direct mapped caches

Systems, apparatuses, and methods for a hardware and software system to automatically decompose a program into multiple parallel threads are described. In some embodiments, the systems and apparatuses execute a method of original code decomposition and/or generated thread execution.

Systems, apparatuses, and methods for a hardware and software system to automatically decompose a program to multiple parallel threads

An embodiment of a semiconductor package apparatus may include technology to identify a nested loop in a set of executable instructions, and determine at runtime if the nested loop is a candidate for cache blocking. Other embodiments are disclosed and claimed.

Post-compile cache blocking analyzer

A processor includes a processor core and a cache controller coupled to the processor core. The cache controller is to allocate, for a memory, a plurality of cache entries in a cache, wherein the processor core is to: detect an amount of the memory installed in a computing system and, responsive to detecting less than a maximum allowable amount of the memory for the computing system, direct the cache controller to increase a number of ways of the cache in which to allocate the plurality of cache entries.

Opportunistic increase of ways in memory-side cache

Techniques are disclosed to identify a frequently-executed region of code during runtime execution of the code, generate initial profiling code for the frequently-executed region of code, cause the initial profiling code to be executed for a minimum number of processing cycles of the computer, and identify replacement candidate store instruction(s) that store a value that is not read by the frequently-executed region of code during execution of the initial profiling code. Replacement candidate load instruction(s) may also be identified that load a value that is not stored or loaded by the frequently-executed region of code during execution of the initial profiling code. Optimized code for the frequently-executed region of code may be generated by replacing each of the replacement candidate store or load instructions(s) with a non-temporal store or load instruction. The optimized code may be executed instead of the frequently-executed region of code during subsequent runtime execution.

Automatic identification and generation of non-temporal store and load operations in a dynamic optimization environment

A processor includes a processing core to execute a transaction with a memory via a cache and a cache controller, associated with the processor, comprising an index mapper circuit to identify a physical memory address associated with the transaction, wherein the physical memory address comprises a plurality of bits, determine, based on the plurality of bits, a first set of bits encoding a tag value, a second set of bits encoding a page index value, and a third set of bits encoding a line index value, determine, based on the tag value, a bit-placement order for combining the second set of bits and the third set of bits, combine, based on the bit-placement order, the second set of bits and the third set of bits to form an index, and generate, based on the index, a mapping from the physical memory address to a cache line index value identifying a cache line in the cache, wherein the processing core is to access, based on the cache line, a memory location referenced by the physical memory address.

In one embodiment, a processor includes: a fetch logic to fetch instructions; a decode logic to decode the instructions; a cache memory; and a control logic to receive a cache filter instruction and responsive to the cache filter instruction enable only a selected portion of a memory address space to be eligible to be cached in the cache memory. The cache filter instruction may indicate the selected portion of the memory address space. Other embodiments are described and claimed.

Instruction and logic for cache control operations

Described herein are mechanisms for continuous automatic tuning of code regions for optimal hardware configurations for the code regions. One mechanism automatically tunes the tunable parameters for a demarcated code region by calculating metrics while executing the code region with different sets of tunable parameters and selecting one of the different sets based on the calculated metrics.

Method, apparatus, system for continuous automatic tuning of code regions

A mechanism for tracking the control flow of instructions in an application and performing one or more optimizations of a processing device, based on the control flow of the instructions in the application, is disclosed. Control flow data is generated to indicate the control flow of blocks of instructions in the application. The control flow data may include annotations that indicate whether optimizations may be performed for different blocks of instructions. The control flow data may also be used to track the execution of the instructions to determine whether an instruction in a block of instructions is assigned to a thread, a process, and/or an execution core of a processor, and to determine whether errors have occurred during the execution of the instructions.

Using control flow data structures to direct and track instruction execution

In one embodiment of the invention, a processor comprising an upper level cache and at least one processor core. The at least one processor core includes one or more registers and a plurality of instruction processing stages: a decode unit to decode an instruction requiring an input of a plurality of data elements, wherein a size of each of the plurality of data elements is less than a cache line size of the processor; an execution unit to load the plurality of data elements to the one or more registers of the processor, without loading data elements spatially adjacent to the plurality of data elements or the plurality of data elements in an upper level cache.

Apparatus and method for considering spatial locality in loading data elements for execution

In an embodiment, a processor includes one or more cores including a first core operable at an operating voltage between a minimum operating voltage and a maximum operating voltage. The processor also includes a power control unit including first logic to enable coupling of ancillary logic to the first core responsive to the operating voltage being less than or equal to a threshold voltage, and to disable the coupling of the ancillary logic to the first core responsive to the operating voltage being greater than the threshold voltage. Other embodiments are described and claimed.

Growing community of inventors

Hillsboro, OR, United States of America

Ruchira Sasanka