From 345c5538f2c5143f1b2f23f861780503437ba22e Mon Sep 17 00:00:00 2001 From: "W. Trevor King" Date: Sat, 13 Nov 2010 10:21:39 -0500 Subject: [PATCH] Add OpenCL link to GPGPU post. --- posts/GPGPU.mdwn | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/posts/GPGPU.mdwn b/posts/GPGPU.mdwn index ef5e8db..2fba683 100644 --- a/posts/GPGPU.mdwn +++ b/posts/GPGPU.mdwn @@ -23,11 +23,11 @@ Specs on NVIDIA's [GeForce GTX 580][]: * 1.5 GB GDDR5 RAM * 192.4 GB/sec memory bandwidth * 1.54 GHz processor clock rate -* 1,581.1 GFLOPs per second +* 1.58 TFLOPs per second Zoom. -The GFLOP/s computaton is `cores⋅clock⋅2`, because (from [page 94][] +The FLOPs/s computaton is `cores⋅clock⋅2`, because (from [page 94][] of the CUDA programming guide) each core can exectute a single multiply-add operation (2 FLOPs) per cycle. Also take a look at the graph of historical performance on [page 14], the table of device @@ -35,6 +35,7 @@ capabilities that starts on [page 111][], and the description of *warps* on [page 93][]. [GPGPU]: http://en.wikipedia.org/wiki/GPGPU +[OpenCL]: http://en.wikipedia.org/wiki/OpenCL [GPU]: http://en.wikipedia.org/wiki/Graphics_processing_unit [NVIDIA]: http://www.nvidia.com/ [CUDA]: http://en.wikipedia.org/wiki/CUDA -- 2.26.2