Bit-Accurate Modeling of GPU Matrix Multiply-Accumulate Units: Demystifying Numerical Discrepancy and Accuracy
This paper proposes closed-loop feature probing (CLFP), a generic and systematic framework for constructing bit-accurate arithmetic behavior models of MMA operations that explain previously observed cross-platform numerical discrepancies and accuracy issues, enable white-box numerical error analysis, and inform softwar...
Peichen Xie, Shuotao Xu, Yang Wang et al.
· 3 citations