Bit-pragmatic Deep Neural Network Computing

TL;DR

This paper introduces Pragmatic (PRA), a novel DNN accelerator architecture that reduces ineffectual computations by calculating only non-zero terms in multiplications, significantly improving performance and energy efficiency.

Contribution

PRA is the first architecture to exploit zero product terms and excess precision in multiplications for DNNs, enhancing efficiency over existing accelerators.

Findings

PRA achieves 2.6x performance improvement over DaDiaNao (DaDN)

PRA improves energy efficiency by up to 28%

Performance gains persist with 8-bit quantization

Abstract

We quantify a source of ineffectual computations when processing the multiplications of the convolutional layers in Deep Neural Networks (DNNs) and propose Pragmatic (PRA), an architecture that exploits it improving performance and energy efficiency. The source of these ineffectual computations is best understood in the context of conventional multipliers which generate internally multiple terms, that is, products of the multiplicand and powers of two, which added together produce the final product [1]. At runtime, many of these terms are zero as they are generated when the multiplicand is combined with the zero-bits of the multiplicator. While conventional bit-parallel multipliers calculate all terms in parallel to reduce individual product latency, PRA calculates only the non-zero terms using a) on-the-fly conversion of the multiplicator representation into an explicit list of powers of two, and b) hybrid bit-parallel multplicand/bit-serial multiplicator processing units. PRA exploits two sources of ineffectual computations: 1) the aforementioned zero product terms which are the result of the lack of explicitness in the multiplicator representation, and 2) the excess in the representation precision used for both multiplicants and multiplicators, e.g., [2]. Measurements demonstrate that for the convolutional layers, a straightforward variant of PRA improves performance by 2.6x over the DaDiaNao (DaDN) accelerator [3] and by 1.4x over STR [4]. Similarly, PRA improves energy efficiency by 28% and 10% on average compared to DaDN and STR. An improved cross lane synchronication scheme boosts performance improvements to 3.1x over DaDN. Finally, Pragmatic benefits persist even with an 8-bit quantized representation [5].