eLife

Revealing global stoichiometry conservation architecture in cells from Raman spectral patterns

Conservation levels of gene expression abundance ratios are globally coordinated in cells, and cellular state changes under such biologically relevant stoichiometric constraints are readable as ...
IEEE

Optimizing Structured-Sparse Matrix Multiplication in RISC-V Vector Processors

Abstract: Structured sparsity has been proposed as an efficient way to prune the complexity of Machine Learning (ML) applications and to simplify the handling of sparse data in hardware. Accelerating ...
IEEE

GRAS: General-purpose and reconfigurable analog solver for matrix computing

Abstract: Matrix computation is ubiquitous in modern scientific and intelligent applications. Conventional digital computing architecture faces fundamental limitations in the post-Moore era, prompting ...
GitHub

llama.cpp — AMD XDNA2 NPU Backend (RyzenAI npu5)

This is a fork of llama.cpp with a custom ggml backend that offloads matrix multiplication to the AMD XDNA2 NPU found in Ryzen AI MAX processors (e.g. Ryzen AI MAX 385). The NPU backend accelerates ...