CUDA Chess » ATI

ATI doesn’t support Radeon 4xxx for OpenCL

iAPX — Thu, 04 Mar 2010 21:07:55 +0000

Looking at the ATI web site and supported OpenCL graphic card, I discovered that any Radeon 4xxxx is marked as “Beta support” for OpenCL: it’s terribly bad to not support the mainstream cards found in the last 2 years on PC (or actual iMac on Windows!).

If you look at nVidia support for OpenCL (& CUDA), every GPU created since 2006 (GeForce 8800 GTS using G80 core) is compatible and fully supported!

Please, ATI, go ahead and ensure full support of OpenCL on these Radeon 4xxx w/ RV700 core (not asking for impossible support of older GPU), don’t let nVidia rule the GPGPU market!

ATI Radeon HD5870 Mobility: OpenCL Teraflop on a laptop

iAPX — Fri, 08 Jan 2010 13:38:52 +0000

The new ATI Radeon HD 5870 use the same 5xxx generation chip that offer full OpenCL 1.0 support, equivalent to nVidia’s GeForce 8xxx and later, but break the Teraflop MAD on a mobile GPU with 1120 Gigaflops!!!

Yes you read this well, you will be able to find laptops with 1 Teraflop raw power inside their GPU before spring, it will be at least 20X the processing power of the multi-core CPU of these laptops!

As a reminder, ASCI RED supercomputer broke the 1 Teraflop barrier in December 1996, with 4510 Pentium-Pro processors. Now you may do calculation at an equivalent pace with a laptop, only 13 years later! ouch!

ATI Radeon 4xxx OpenCL benchmarks

iAPX — Wed, 11 Nov 2009 02:18:42 +0000

There’s some OpenCL benchmarks out there, and on OpenCL Benchmark that test real GPGPU computation, instead of pure processing power on theorical computation, Radeon 4xxx series lag far far behind of nVidia actual GPU.

ATI Radeon 3xxx and 2xxx are not supported, albeit nVidia’s GPU are supported since 2006 G80 (GeForce 8800 and any GeForce 8 series or later GPU), and Radeon 4xxx are just underperforming, lacking shared memory (memory inside each processor core).

Lacking “shared memory” means that for any data access Radeon 4xxx have to access global video card memory, that is usually 20X to 30X slower, and worse, memory bandwidth on Radeon graphic card are 2X to 3X slower than on nVidia’s. This is not an handicap for games, where radeon are really great graphic card, but it is for GPGPU and OpcnCL.

The result of lacking Shared Memory and slow graphic card memory: a Radeon 4870 (around 200$ street) could not compete with GeForce 9400M IGP (found on Mac Mini, MacBook Air, MacBook…), and a GeForce 9400M iMac will beat any ATI Radeon 4850 iMac when it’s time to compare OpenCL performances!

ATI’s OpenCL CPU-Only!

iAPX — Fri, 07 Aug 2009 14:06:39 +0000

While nVidia actually support OpenCL on it’s GPU, but not on main CPU, ATI offers it’s own drivers that support main CPU but not it’s GPU! Anyway ATI’s GPU are not really ceonceived for GPGPU and wil llag far far behind nVidia’s on real OpenCL implementations!

The purpose of OpenCL is to enable code to run on both CPU and GPUs (even a mix of ATI and nVidia), not to enable to run either in CPU (what’s the novelty???) or restricted to a propretary GPU!!!

At this time, CUDA seems to be the technology path to follow before switching to OpenCL in 2010 or 2011…