Hybrid SLC-MLC RRAM Mixed-Signal Processing-in-Memory Architecture for Transformer Acceleration via Gradient Redistribution

TL;DR

This paper introduces HyFlexPIM, a hybrid RRAM-based processing-in-memory architecture for transformer acceleration that combines SLC and MLC technologies with a gradient redistribution algorithm to optimize accuracy and efficiency.

Contribution

The paper presents a novel hybrid SLC-MLC RRAM PIM architecture with a gradient redistribution algorithm for optimized transformer inference acceleration.

Findings

Achieves up to 1.86X higher throughput.

Achieves up to 1.45X higher energy efficiency.

Effectively balances accuracy and efficiency using hybrid RRAM and gradient redistribution.

Abstract

Transformers, while revolutionary, face challenges due to their demanding computational cost and large data movement. To address this, we propose HyFlexPIM, a novel mixed-signal processing-in-memory (PIM) accelerator for inference that flexibly utilizes both single-level cell (SLC) and multi-level cell (MLC) RRAM technologies to trade-off accuracy and efficiency. HyFlexPIM achieves efficient dual-mode operation by utilizing digital PIM for high-precision and write-intensive operations while analog PIM for high parallel and low-precision computations. The analog PIM further distributes tasks between SLC and MLC PIM operations, where a single analog PIM module can be reconfigured to switch between two operations (SLC/MLC) with minimal overhead (<1% for area & energy). Critical weights are allocated to SLC RRAM for high accuracy, while less critical weights are assigned to MLC RRAM to maximize capacity, power, and latency efficiency. However, despite employing such a hybrid mechanism, brute-force mapping on hardware fails to deliver significant benefits due to the limited proportion of weights accelerated by the MLC and the noticeable degradation in accuracy. To maximize the potential of our hybrid hardware architecture, we propose an algorithm co-optimization technique, called gradient redistribution, which uses Singular Value Decomposition (SVD) to decompose and truncate matrices based on their importance, then fine-tune them to concentrate significance into a small subset of weights. By doing so, only 5-10% of the weights have dominantly large gradients, making it favorable for HyFlexPIM by minimizing the use of expensive SLC RRAM while maximizing the efficient MLC RRAM. Our evaluation shows that HyFlexPIM significantly enhances computational throughput and energy efficiency, achieving maximum 1.86X and 1.45X higher than state-of-the-art methods.