Better OpenCL support in NVIDIA's CUDA SDK

[LtMatt]

Introduction: CUDA, OpenCL, GPGPU

CUDA and OpenCL are techniques to program GPUs, where CUDA only works on NVIDIA GPUs. CUDA and OpenCL are often referred to as GPGPU- General-purpose computing on graphics processing units. Various researchers have concluded that with OpenCL and CUDA comparable results can be reached.
Besides GPUs, OpenCL is designed to also work on DSPs, mobile processors, FPGAs and modern CPUs. It is not performance-portable, but all techniques learnt can be used on the various platforms.
OpenCL is very promising, but like with all open standards it needs support of the developers and goodwill from the participating companies.


The problem

Nvidia is not including OpenCL samples in the latest CUDA SDK and has removed profiler-support for OpenCL, and instead focusing more on their proprietary CUDA. The reason is simple: every developer who chooses CUDA over OpenCL, is limited to NVIDIA hardware.
As a Khronos member with an excellent record in implementing and promoting standards like OpenGL, this is a surprising and even unacceptable behavior from Nvidia.
OpenCL developers need a full-blown SDK (such as OpenCL samples and aprofiler), so the potential and limitations of NVIDIA GPUs can be learned. Also industry standards like OpenCL help in building up a bigger market for GPU computing, and will be beneficial to Nvidia in the long term.
What I aim by signing this petition
By signing this petition, I request Nvidia to put back the OpenCL samples and profiler in their latest CUDA SDK. By this, the choice is put back to the developer, as it should.

If you would like to sign please visit - http://www.ipetitions.com/petition/opencl-examples-in-cuda-5-sdk/

 

[LtMatt]

Its going to benefit us all.

Signed.

 

[LtMatt]

It will only benefit those with an ATI card and why should NVidia help ATI?

Just looked at your sig(ATI card)

Read the OP again.

 

[LtMatt]

I give up.

 

[LtMatt]

I can't quite tell if that's sarcasm

It is.

 

[LtMatt]

That's an old benchmark result, things have changed a lot since then for both sides. A lot of benches tend to disable TressFX as well for some reason which always baffles me as its not like Physx where one side is locked out. All that does though is allow the Nvida cards (excluding the titan and to a lesser degree the 780) to score more fps.

 

[LtMatt]

Its not in there I can't remember the thread. Matt?

Yes i remember what he's talking about. Can't remember what thread it was in though sorry. It was very close to a 780 though few fps behind from memory.

 

[LtMatt]

I thought I'd do a quick test. I copied the res and settings from here:

Spoiler

My results with identical settings:

Spoiler

My Specs (SLI was disabled for the test):

Spoiler

As you can see, a mildly overclocked Titan can beat a heavily overclocked 7970 at a title which AMD has every advantage in. So you can understand why I say Nvidia's OpenCL game performance is not quite the apocalypse you think it is.

Sorry for all the OT.

Thanks for posting but why on earth are you comparing a £900 gpu vs a £300 gpu? (actually it was £270 after i sold the games)

I think all this does is emphasise what a kick ass card a clocked 7970 ghz is.

 

[LtMatt]

TR has to be pretty much the current worst case scenario for NV cards as well.

Now back to our scheduled petty arguing and name calling.

No that's probably Hitman Resolution, closely followed Sleeping Dogs, Tomb Raider comes after that.

Regardless i don't think a 13% advantage on a card costing 200% less is anything to get excited about, with all due respect.