Reproduce “Superscalar Loop Processors” on OpenResearch

Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.

Universal app for Apple silicon and Intel. Requires macOS 11 or later.

About this paper

Luca Colagrande, Pascal Etterli, Luca Benini

Schnizo builds a reusable dependency schedule for static loops, recording instruction relationships during an initial single-issue pass and replaying loop iterations with independent functional-unit clusters issuing in parallel. In simulation, specialized Schnizo-LA achieves a 4.4× geometric-mean speedup on AXPY and DOT and 72% higher core area efficiency than scalar execution. These results are limited to the tested kernels and configuration.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.