Unlimited Vector Extension with data streaming support

João Mário Ribeiro Domingos

Key information

Authors:

João Mário Ribeiro Domingos (João Mário Ribeiro Domingos)

Supervisors:

Nuno Filipe Valentim Roma (Nuno Filipe Valentim Roma); Pedro Filipe Zeferino Tomás (Pedro Filipe Zeferino Tomás)

Published in

October 13, 2020

Abstract

Unlimited vector extension (UVE) is a novel instruction set architecture extension that takes streaming and SIMD processing together into the modern computing scenario. It aims to overcome the shortcomings of state-of-the-art scalable vector extensions by adding data streaming as a way to simultaneously reduce the overheads associated with loop control and memory access indexations, and memory access latency. This is achieved through a set of instructions which are able to pre-configure the loop memory access pattern(s), attaining accurate and timely data prefetching on predictable access patterns, such as in multidimensional arrays or on indirect memory access patterns. Each of the configured data streams is associated with a general vector register, which is then used to interface with the streams. In particular, iterating over the stream is simply achieved by reading/writing to the corresponding input/output stream, as the data is instantly consumed/produced. To evaluate the proposed UVE, a first gem5 implementation was made on an out-of-order processor model, based on the ARM Cortex-A76, thus taking into consideration typical speculative and out-of-order execution found in high-performance computing processors. The evaluation was carried out on a set of representative kernels, by assessing the number of executed instructions, its impact on the memory bus and its overall performance. Compared with state-of-the-art solutions, such as the upcoming ARM Scalable Vector Extension (SVE), results show that the proposed solution attains performance speedups between 2x and 4x.