• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

Better OpenCL support in NVIDIA's CUDA SDK

Caporegime
Joined
12 Jul 2007
Posts
43,716
Location
United Kingdom
ohtov5.jpg


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/
 
A bit of a bizarre one this, at first I was thinking why should they support a competing standard in their own proprietary one. It would be like apple promoting android apps and vice a versa.
But then after rereading it I see they used too but have now decided to remove it, which is quite strange, so I agree, bad form Nvidia.

Signed.
 
Read the OP again. ;)

As I said the only people who want this is ATI people. All good 3D\Photo\Video software uses Cuda

As we all know hardware is faster then software. Why not ask ATI to get OpenCL working like Cuda does?

And as for "open standards are better for everyone" did that guy ever work in the network trade when it first started?

This just AMD fan boys wanting something for nothing AGAIN.
 
Why is this a surprise to anyone?

If the tables were turned and AMD had got a foothold on the GPGPU market first do you think they would be the stalwarts of openness they are now? Unlikely.

End user still has a choice what GPU they want to buy and which implementation of compute they want. It's not like nVidia has a monopoly on the compute world, they just have a far bigger market with their compute cards....because they pushed for it prior to AMD/ATI did. Why should they weaken their position by competing against their own CUDA technology? It's bad business sense, no matter what is supposedly 'better' for the end user.

Want OpenCL? Buy AMD
Want CUDA? Buy nVidia

/endof
 
Last edited:
OpenCL (Open Computing Language) was developed by Apple and they still hold Trademark rights. Intel, AMD, Nvidia, Altera, Samsung, Vivante and ARM Holdings use OpenCL.

Cuda (Compute Unified Device Architecture) was developed by Nvidia and used by Nvidia.

Do AMD actually develop anything?
 
Regardless, to not support an Open standard seems silly.

On the otherhand, by not supporting it, they're prevent OpenCL from gaining a foothold, in doing that however they're also preventing open progression, as it'll be strictly tied to Nvidia.

However, with consoles being AMD, consoles being the defacto platform, surely OpenCL will become a standard based on that? Unless Nvidia have allowed GPU accelerated PhysX to work on both consoles?

Certainly not a black/white.
 
Do AMD actually develop anything?

lara's hair? :D

Regardless, to not support an Open standard seems silly.

On the otherhand, by not supporting it, they're prevent OpenCL from gaining a foothold, in doing that however they're also preventing open progression, as it'll be strictly tied to Nvidia.

However, with consoles being AMD, consoles being the defacto platform, surely OpenCL will become a standard based on that? Unless Nvidia have allowed GPU accerlated PhysX to work on both consoles?

they do support it, it still runs on nvidia hardware, though they removed strong hardware support for it, having been criticised for big chunks of Fermi being underutilised, all they've done is removed code samples from CUDA development kits, which is hardly surprising
based on what little info we have already - my guess would be that Maxwell is going to have much stronger OpenCL hardware support

Nvidia have indeed allowed physx to work on both consoles :D
 
Last edited:
Not hardware accelerated PhysX they haven't, not that the 360/PS3 could handle it.
Although I've probably got the feeling this has been made out to be something that it's not :p? Which wouldn't surprise me.
 
Regardless, to not support an Open standard seems silly.

On the otherhand, by not supporting it, they're prevent OpenCL from gaining a foothold, in doing that however they're also preventing open progression, as it'll be strictly tied to Nvidia.

However, with consoles being AMD, consoles being the defacto platform, surely OpenCL will become a standard based on that? Unless Nvidia have allowed GPU accelerated PhysX to work on both consoles?

Certainly not a black/white.

lara's hair? :D

It could just be a fact that Nvidia are feeling put out with not adding GPU's to the latest consoles and spitting their dummy out.

I was reading through a couple of forums on PhysX yesterday and the usual Nvidia Vs AMD was in full swing but one comment sat with me. The guy said "All I hear is AMD are much cheaper and provide the better bang for buck performance, I paid more for my Nvidia card than the comparable AMD card, so why should they have PhysX? I paid extra for that privilege" (or something very similar) but it did make me stop and think for a bit. I am not trying to justify that comment but in fairness, there is truth to what he says.
 
Back
Top Bottom