摘要:
Various embodiments of the invention concern methods and apparatuses for power and time efficient load handling. A compiler may identify producer loads, consumer reuse loads, consumer forwarded loads, and producer/consumer hybrid loads. Based on this identification, performance of the load may be efficiently directed to a load value buffer, store buffer, data cache, or elsewhere. Consequently, accesses to cache are reduced, through direct loading from load value buffers and store buffers, thereby efficiently processing the loads.
摘要:
Various embodiments of the invention concern methods and apparatuses for power and time efficient load handling. A compiler may identify producer loads, consumer reuse loads, consumer forwarded loads, and producer/consumer hybrid loads. Based on this identification, performance of the load may be efficiently directed to a load value buffer, store buffer, data cache, or elsewhere. Consequently, accesses to cache are reduced, through direct loading from load value buffers and store buffers, thereby efficiently processing the loads.
摘要:
In one embodiment, the present invention includes a software-controlled method of forming instruction strands. The software may include instructions to obtain code of a superblock including a plurality of basic blocks, build a dependency directed acyclic graph (DAG) for the code, sort nodes coupled by edges of the dependency DAG into a topological order, form strands from the nodes based on hardware constraints, rule constraints, and scheduling constraints, and generate executable code for the strands and store the executable code in a storage. Other embodiments are described and claimed.
摘要:
In one embodiment, the present invention includes a software-controlled method of forming instruction strands. The software may include instructions to obtain code of a superblock including a plurality of basic blocks, build a dependency directed acyclic graph (DAG) for the code, sort nodes coupled by edges of the dependency DAG into a topological order, form strands from the nodes based on hardware constraints, rule constraints, and scheduling constraints, and generate executable code for the strands and store the executable code in a storage. Other embodiments are described and claimed.
摘要:
In one embodiment, the present invention introduces a speculation engine to parallelize serial instructions by creating separate threads from the serial instructions and inserting processor instructions to set a synchronization bit before a dependence source and to clear the synchronization bit after a dependence source, where the synchronization bit is designed to stall a dependence sink from a thread running on a separate core. Other embodiments are described and claimed.
摘要:
An apparatus and method is described herein for conditionally committing and/or speculative checkpointing transactions, which potentially results in dynamic resizing of transactions. During dynamic optimization of binary code, transactions are inserted to provide memory ordering safeguards, which enables a dynamic optimizer to more aggressively optimize code. And the conditional commit enables efficient execution of the dynamic optimization code, while attempting to prevent transactions from running out of hardware resources. While the speculative checkpoints enable quick and efficient recovery upon abort of a transaction. Processor hardware is adapted to support dynamic resizing of the transactions, such as including decoders that recognize a conditional commit instruction, a speculative checkpoint instruction, or both. And processor hardware is further adapted to perform operations to support conditional commit or speculative checkpointing in response to decoding such instructions.
摘要:
An apparatus and method is described herein for coupling a processor core of a first type with a co-designed core of a second type. Execution of program code on the first core is monitored and hot sections of the program code are identified. Those hot sections are optimize for execution on the co-designed core, such that upon subsequently encountering those hot sections, the optimized hot sections are executed on the co-designed core. When the co-designed core is executing optimized hot code, the first processor core may be in a low-power state to save power or executing other code in parallel. Furthermore, multiple threads of cold code may be pipelined on the first core, while multiple threads of hot code are pipeline on the co-designed core to achieve maximum performance.
摘要:
Technologies for generating composable library functions include a first computing device that includes a library compiler configured to compile a composable library and second computing device that includes an application compiler configured to compose library functions of the composable library based on a plurality of abstractions written at different levels of abstractions. For example, the abstractions may include an algorithm abstraction at a high level, a blocked-algorithm abstraction at medium level, and a region-based code abstraction at a low level. Other embodiments are described and claimed herein.
摘要:
A system is disclosed that includes a processor and a dynamic random access memory (DRAM). The processor includes a hybrid transactional memory (HyTM) that includes hardware transactional memory (HTM), and a program debugger to replay a program that includes an HTM instruction and that has been executed has been executed using the HyTM. The program debugger includes a software emulator that is to replay the HTM instruction by emulation of the HTM. Other embodiments are disclosed and claimed.
摘要:
Logic and instruction to efficiently monitor loop trip count. Loop trip count information of a loop may be stored in a dedicated hardware buffer. Average loop trip count of the loop may be calculated based on the stored loop trip count information. Based on the average trip count, loop optimizations may be applied or removed from the loop. The stored loop trip count information may include an identifier identifying the loop, a total loop trip count of the loop, and an exit count of the loop.