Hello out there. Just read this a day ago at Phoronix: http://www.phoronix.com/scan.php?page=news_item&px=OTg0MQ I've also read that there is work done on the PTX assembly backend in LLVM for generation code for nVidia GPUs. I'm still grasping for the silky fathem having GPGPU on FreeBSD anyway like CUDA or OpenCL as the Linux and Windows fellows already have as well as the guys from OS X. Since nVidia offers 64bit ready BLOBs for their nice GPUs, I still hope that with LLVM, and maybe a proper OpenCL frontend like CLANG, FreeBSD users will have the chance to execute OpenCL code on a GPU. We use this stuff in science for now (and its done exclusiveley on Linux so far, since we use the CUDA SDK to generate OpenCL 1.1 code executed on TESLA and GTX570 graphics boards which give us impressive performance for our astrodynamical modelling ...). Sorry, if someone feels bothered by my repetitive bringing up this subject ... I still does not have lost all hope for FreeBSD. ohReceived on Mon Aug 29 2011 - 08:13:21 UTC
This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:40:17 UTC