Introduction to Dynamic Code Generation an Experiment with Matrix Multiplication for the STHORM Platform
It’s sometimes fun to re-read our old publication (from 2013). This publication was an experimentation of dynamic code generation on the STHORM platform aka P2012 (somebody remembers this nice parallel MPSoC ? :-)
The experimentation was about dynamic code specialization and dynamic unrolling for a simple matrix multiplication kernel
This article was publishd here : https://link.springer.com/chapter/10.1007%2F978-1-4614-8800-2_6 and is now available here : https://hal.archives-ouvertes.fr/view/index/docid/3437284 and also inline on this web page :