3 ms·

While I'm excited to see AMD's offering, as a scientific-HPC user I can't help but wonder how much marketshare AMD will be able to gain without more information

While I'm excited to see AMD's offering, as a scientific-HPC user I can't help but wonder how much marketshare AMD will be able to gain without more information on supporting software - specifically good compilers + math libraries (cf. Intel compilers + MKL). Strangely, I've not seen much on HN, or elsewhere, make mention of AMD's software support. Is this because it doesn't exist, or because compilers are less "sexy" than shiny new hardware?

AlphaSite 9y ago
They have a custom fork of Clang http://developer.amd.com/amd-optimizing-cc-compiler-aocc-technical-support/ http://developer.amd.com/amd-optimizing-cc-compiler-aocc-tec... for C and ROCm for GPUs https://github.com/RadeonOpenCompute/ROCm https://github.com/RadeonOpenCompute/ROCm
geezerjay 9y ago
> While I'm excited to see AMD's offering, as a scientific-HPC user I can't help but wonder how much marketshare AMD will be able to gain without more information on supporting software - specifically good compilers + math libraries (cf. Intel compilers + MKL). My take is that AMD's newest offering will be very well received by everyone who has a relatively tight budget but needs a small supercomputer on the desktop. This means data analysts and people doing all kinds of structural analysis work. As some optimization algorithms fit the definition of embarrassingly parallel, the expected turn-around time of anyone doing that sort of work will benefit greatly from the extra speed, bandwidth and core count of AMD's Ryzen/Threadripper/Epyc line.