English

IM-Unpack: Training and Inference with Arbitrarily Low Precision Integers

Machine Learning 2024-03-13 v1 Computation and Language Computer Vision and Pattern Recognition

Abstract

GEneral Matrix Multiply (GEMM) is a central operation in deep learning and corresponds to the largest chunk of the compute footprint. Therefore, improving its efficiency is an active topic of ongoing research. A popular strategy is the use of low bit-width integers to approximate the original entries in a matrix. This allows efficiency gains, but often requires sophisticated techniques to control the rounding error incurred. In this work, we first verify/check that when the low bit-width restriction is removed, for a variety of Transformer-based models, whether integers are sufficient for all GEMMs need -- for {\em both} training and inference stages, and can achieve parity with floating point counterparts. No sophisticated techniques are needed. We find that while a large majority of entries in matrices (encountered in such models) can be easily represented by {\em low} bit-width integers, the existence of a few heavy hitter entries make it difficult to achieve efficiency gains via the exclusive use of low bit-width GEMMs alone. To address this issue, we develop a simple algorithm, Integer Matrix Unpacking (IM-Unpack), to {\em unpack} a matrix with large integer entries into a larger matrix whose entries all lie within the representable range of arbitrarily low bit-width integers. This allows {\em equivalence} with the original GEMM, i.e., the exact result can be obtained using purely low bit-width integer GEMMs. This comes at the cost of additional operations -- we show that for many popular models, this overhead is quite small.

Keywords

Cite

@article{arxiv.2403.07339,
  title  = {IM-Unpack: Training and Inference with Arbitrarily Low Precision Integers},
  author = {Zhanpeng Zeng and Karthikeyan Sankaralingam and Vikas Singh},
  journal= {arXiv preprint arXiv:2403.07339},
  year   = {2024}
}
R2 v1 2026-06-28T15:16:45.398Z