Is OpenCL 1.2 on it's way?

Can anybody shed some light on Nvidia’s future OpenCL plans? They were quick to support OpenCL 1.0 and 1.1 but lately it has gotten pretty quite…

I know that nothing has been announced nevertheless I’d like to know if I should even bother spending time on OpenCL 1.2. AMD and Intel support it and now only Nvidia is missing.

[edit]Even though I just came across https://devtalk.nvidia.com/default/topic/522714/cuda-programming-and-performance/opencl-1-2/ the question remains the same.[/edit]

I doubt that NVIDIA would do so. Although, I would very much suggest getting up to speed with OpenCL 1.2 nevertheless.

1.2 is cool - with all the new sub-buffer support - One can do multi-GPU with so much ease.
Easy to write libraries with m-gpu support as well.
Buffer migration is also a worthwhile thing to do.

I hope NVIDIA follows it up with a 1.2 release. Otherwise, serious developers might start migrating to other platforms. NV would definitely not like it.

I am building my career in parallel programming and not being able to use OpenCL 1.2 yet is definitely not good for my progress. I understand the financial reasons but honestly, CUDA 5 can still top OpenCL 1.2 in performance so I really don’t understand why they cannot give us new drivers for OpenCL 1.2 when even the CPU can use OpenCL nowadays. come on NVIDIA, I love you folks but you’re killing me :(

Nvidia ignoring is customer base is a good way to lose out on money

IMHO, This link describes OpenCL status for discrete/hpc:

nVidia and Intel ignoring OpenCL for the same reason.

If it is a bother for you guys, vote with your wallet, return the card(s) and communicate to the vendor and NVIDIA explicitly the reason you did so was because of lack of OpenCL 1.2 support.

I’m already voting with my wallet. Nvidia will not receive a single cent from me until

  • OpenCL 1.2 support
  • Less crippling of DP compute performance for high end parts (GTX780(m), …)

Having much lower OpenCL compute performance than AMD per $ and then crippling it even further is a really bad joke (hate goes out to nv marketing / product management guys).

Nvidia can count itself lucky that it was able to convince so many OEMs to pick Nvidia discrete graphics.

AMD: OpenCL 1.2 Support
Intel: OpenCL 1.2 Support
Nvidia: “Industry Leading” OpenCL 1.1 Support

Qualcomm will probably have working OpenCL 1.2 Support on mobile devices before Nvidia on their GPUs. Sad.

Back in the days when I started OpenCL I chose it because I said to myself: hey, this rocks. Royalty free, it only grasps the core of parallel programming, CPU/GPU with one rock… it’s cool.

A few years pass, and basically I’m back where I started, my codebase is not portable (or I have to make workarounds if at all possible), because NV is sabotaging their own implementation. Only one thing changed, now my codebase is stuck with AMD (and Intel). To be honest, it’s a fair exchange. The compute power I get for my (company’s) money is far superior then if I went with NV.

Yes, there are the CUDA features that rock, but we all know it is because of ignorant and incompetent market strategy that they are not implemented in OpenCL. That is what vendor-specific extensions are for. RDMA, dynamic parallelism… all the goodness could be imported via extensions, but nooooo. You should buy Tesla and use CUDA. Uhm… why?

No wonder PS4 and XBox One went with AMD instead, who took the chance and helped develop all the tools needed for a proper heterogenous platform. And look at all the ingame trailers: they are jawdropping. On the other hand NV has their own complete platform in the form of Tegra 4. Fixed function shader HW in 2013??? Are you joking??? Yet another kick into developers who have to find workarounds if they want cool compute features in their apps on GPUs (facial recognition and the likes). Because of NV, their code is yet again not portable.

And the biggest joke is, that while it is completely clear that the tool/driver/SDK developing capacity of NV is far superior than any other company’s on the market, they don’t even have 1.2 conforming OpenCL implementation, while AMD for one has the conforming SDK, driver and tools THE NEXT DAY the spec is released. (Intel came 1-2 months later) Plus, AMD and Intel have already started building higher-level tools on top of it, like Shevlin Park or Bolt, which use OpenCL as a compute backend.

NV is doing practically nothing and that’s exactly why it’s getting nowhere.

NV going nowhere??? wow …

OpenCL is dead (at least for discrete). It was never that big of a find.
If you’ll check closer (probably when this month’s Top500 list comes with China’s MIC #1) you’ll find that soon the OpenCL will be useful for you only on AMD’s platforms (no dynamic parallelism??
what about X-less apps running on Linux not via root permission? stability? FFT/Blas/…?)
Intel will go with MKL and Cilk leaving OpenCL as a step-son, much like NVidia did.

What’s the point of OpenCL if it will all be filled with extensions? its not performance
cross-platform and it is not feature cross-platform.

Mobile will follow this trend - Imagination for example seems to push RenderScript and not
OpenCL. NVidia will have CUDA of course and other companies might find that OpenCL is not
that great either (Qualcomm already have their extensions and OpenCL on mobile is only what?
a year old???).

At the end the hardware’s are mostly similar (at least in performance) so the main
different for the vendors would be in the eco-system, software and libraries.
And we all know, as you said, who is the leader in those fields. :)

I think the situation with OpenCL will sort itself out eventually. Right now the fundamental architectures of GPU-based accelerators are changing so fast, and so completely different between manufacturers, that I can see how standardization of a programming interface might not be a high priority for the some of the players. Once the major changes slow down a bit, I think we’ll see more effort put into interoperability…