We've released SETI@home 7.03 for Intel GPUs that support OpenCL 1.2. If you have an Intel GPU and want to use it to speed up SETI@home make sure that "Use Intel GPU" is set to "yes" in your Project Preferences.

Thanks again to the crew at lunatics.kwsn.net for porting and testing this app.

---

@SETIEric@qoto.org (Mastodon)

How do we know what Intel GPU's support the OpenCL 1.2?

---

You can find a list of all Intel GPUs here along with their corresponding Intel CPU model numbers they are integrated into (if applicable). Intel's seventh generation of GPUs all support OpenCL 1.2.

Basically if you have a 3rd generation Core i3/i5/i7 processor or later, your GPU should support OpenCL 1.2.

One question: I use a i7 4770k with nvidea gtx 770 win 7 and boinc version 7.2.33 if i activate use of intel gpu did the system use both gpus, the nvidea and the intel ?

One question: I use a i7 4770k with nvidea gtx 770 win 7 and boinc version 7.2.33 if i activate use of intel gpu did the system use both gpus, the nvidea and the intel ?

Boinc doesn't detect your Intel GPU at the moment, you may need a use a Dummy plug, or extend the Desktop, and then force it to use VGA to get it to stay detected:

Computer 7062032:

NVIDIA GeForce GTX 770 (2048MB) driver: 332.21 OpenCL: 1.01

That is right click on the desktop, chose 'Screen resolution', click the Detect Button, then find the Intel GPU, and where it says Multiple displays 'No display detected' change it to 'Try and connect anyway on:VGA',
then restart Boinc.

Claggy

Thanks for the heads-up, Eric, and congrats for this new feature.
It's cool for all the owners of Intel GPU.
Thanks very much to all the LUNATICS crew who are always on the breach :)

---

Just be aware that using this with a stock Intel HSF or insufficient airflow may cause overheating problems (especially with Haswell based CPU's).

Cheers.

Hi there,

I am using the lunatics optimised apps at the moment (mostly due to an incompatibility between the standard nvidia apps and my GTX660).

Just wondering if anyone can suggest an app_info.xml snippet to enable work fetch of Intel GPU work units on an anonymous platform.
I expect I will have to download the application manually too.

I have added <use_all_gpus>1</use_all_gpus> to my cc_config.xml file and the BOINC client shows that it has detected the Intel GPU
OpenCL CPU: Intel(R) Core(TM) i7-3770K CPU @ 3.50GHz (OpenCL driver vendor: Intel(R) Corporation, driver version 3.0.1.10878, device version OpenCL 1.2 (Build 76413))
My project settings allow all GPU types.

I will be interested to see if the project favours one type of GPU for processing or if it keeps work queued for all available GPUs.

Regards,
Daniel

---

here: http://setiathome.berkeley.edu/forum_thread.php?id=71867

---

SETI apps news
We're not gonna fight them. We're gonna transcend them.

Question regarding the GPU acceleration. On an nVidia card is CUDA still used or is OpenCL used on all cards now? I was also wondering if SETI@home uses SP or DP for the calculation or if there is chance to optimize either way.

---

Quick, semi-related question: Is there an easy way to tell if your system is using the GPU or not?

I've found places to turn using the GPU on or off, but haven't found anyplace where I can check the running process to make sure SETI is really using the GPU.

---

Question regarding the GPU acceleration. On an nVidia card is CUDA still used or is OpenCL used on all cards now?

I believe nVidia is focusing exclusively on their proprietary CUDA initiative, so CUDA apps will be required. OpenCL support will depend on if nVidia stops supporting the OpenCL API library or specifically removes hardware support.

I was also wondering if SETI@home uses SP or DP for the calculation or if there is chance to optimize either way.

SETI@home uses single precision floats.

Quick, semi-related question: Is there an easy way to tell if your system is using the GPU or not?

There are a couple ways you can tell.

1) When you open BOINC Manager, you should see on the Tasks page that a particular workunit is using something like "0.25 CPUs and 1 GPU".

2) If you're familiar with the executable names, you can open Task Manager and look at the running processes for something stating nvidia/ati/cuda/opencl.

Quick, semi-related question: Is there an easy way to tell if your system is using the GPU or not?

I've found places to turn using the GPU on or off, but haven't found anyplace where I can check the running process to make sure SETI is really using the GPU.

You can check your Boinc startup in the Event Log to see if a GPU is detected,
you can check your list of computers to see if Boinc reports a GPU to the server,
You can check your list of tasks to see if any have been allocated to a GPU app,
In your case not, as a third generation Core CPU (ie an Intel(R) Core(TM) I*-3*** or later) is required for an Intel OpenCL Capable GPU to be available:

Computers belonging to seybernetx

Claggy

Question regarding the GPU acceleration. On an nVidia card is CUDA still used or is OpenCL used on all cards now? I was also wondering if SETI@home uses SP or DP for the calculation or if there is chance to optimize either way.

The apps marked as Cuda are Cuda apps, the apps marked as OpenCL are OpenCL apps, the app marked as cal is a Brook+ app:

Setiathome Applications

Any SP or DP Cuda capable GPU can be used for the Cuda and Nvidia OpenCL apps, any AMD/ATI HD5*** and later GPU can be used for the AMD/ATI OpenCL apps, HD4*** GPUs are limited to the AMD/ATI OpenCL Astropulse app,

Just because SP can be used doesn't mean there is only SP code, there are some different code paths for different GPUs, meaning DP code could be used where available.

Claggy

Does That mean that Mac's with Intel GPU's will be able to partake in GPU Crunching?

Thanks

---

Does That mean that Mac's with Intel GPU's will be able to partake in GPU Crunching?

Thanks

Only windows app for now. Maybe our Linux guru will provide MacOS port later, lets hope for that.

---

SETI apps news
We're not gonna fight them. We're gonna transcend them.

Thanks for the update

---

Only windows app for now. Maybe our Linux guru will provide MacOS port later, lets hope for that.

So just to confirm, this isn't supported on Linux yet? Is there a timeline for a Linux Intel OpenCL app?

Only windows app for now. Maybe our Linux guru will provide MacOS port later, lets hope for that.

So just to confirm, this isn't supported on Linux yet? Is there a timeline for a Linux Intel OpenCL app?

Nope as far as I know there are no plans for any Linux Intel OpenCL GPU apps,

Before that can happen Boinc needs to be able to detect the Intel GPU, and for that to happen Intel needs to support OpenCL on Intel GPUs on Linux,
or the beignet project to enable Intel OpenCL GPU support:

http://software.intel.com/en-us/forums/topic/402118

Right, the XE version of the SDK supports Linux for CPUs and Intel Xeon Phi Coprocessor. Our product support matrix is expanding all the time with just recently a new support on Windows for 4th Generation Intel Core Processor with Intel Iris Graphics.

We will use this forum to update you when we will have more news on new products and OSs.

Meanwhile, you may want to try the open source project beignet. This project provide am open source alpha software for OpenCL on Intel HD Graphics 4000. It is not commercial product by Intel, but is available for the open source community.

Arnon

and a reason why Intel may never support Intel GPUs on Linux:

Intel will *likely never* support linux OpenCL on its Core processors. You just to have to follow the gigaflops to understand why...

Intel sell a *lot* of Xeons into supercomputing centers, where Xeons are bought hundreds, sometimes thousands at a time. These centers tend to run OpenCL on linux. Which is great...only not. If you look at a Xeon that was released in any given quarter, and then look at a Core released the same quarter, the Xeon tends to be $100-200 more expensive. Which is fine - it can pull more gflops, right? No, not on OpenCL (with a driver). Go look at that Xeon again, its gigaflops. Now look at the Core...but this time, look at the glops on the GPU. Add that to the gflops from the CPU. The Core has more total gigaflops, right? OpenCL doesn't care if the CPU or the GPU is the thread - that's the point of heterogenous computing. And the point of why Intel won't support it on linux.

So, if Intel were to support Core CPU/iGPU's on linux, somebody would eventually figure out that they can save thousands of dollars, and get more performance for their 1000-node supercomputer if they built it on Core. So, don't expect to see a linux Core OpenCL driver any time soon...unfortunately Intel will protect their interests on this one.

Claggy