Jump to content

Featured Replies

  • Replies 731
  • Views 125.6k
  • Created
  • Last Reply

Top Posters In This Topic

Most Popular Posts

  • Fair warning: GPUPI 3.3 is now officially available and will give a decent speedup on all calculations. I'd like to clarify the reasons behind this, because I know that means some rebenching might be

  • Leeghoofd
    Leeghoofd

    I already talked to Matt today and this version will probably be called GPUPi4 and we will have to decide on other benchmarks for it.    Let it be clear that I will not let this benchmark ve

  • This sounds ok. But when will the new one get points rev 8? I may be in the grave when that is released.  In regards to mat i hope he doesnt take this as disrespect by hwbot. He has made probably

Posted Images

  • Author

Parallel computations like in GPUPI can not be validated during a run. The result is an accumulation of billions of smaller partial results, that have been crunched by each core.

Pi Bench : The Return of the NotExactInRound

 

I don't think you can compare this "GPUI bench" to super pi,they are in no way the same imo,maybe gpui is comparable to hwbot prime on software requirements

Edited by Bullant

I don't think you can compare this "GPUI bench" to super pi,they are in no way the same imo,maybe gpui is comparable to hwbot prime on software requirements

 

I don't that's completely fair, HWBot prime is the only benchmark I know where the real stress testing begins when you save the screenshot ;)

I don't that's completely fair, HWBot prime is the only benchmark I know where the real stress testing begins when you save the screenshot ;)

 

Ahah,I was just meaning that gpui efficiency seems to come from software vs super Pi is different,I'm not bagging the bench as every bench requires the tweaker to get the best efficiency

  • Author

Not true. GPUPI is just not memory bandwidth limited especially on GPUs where memory is already very fast. It's like SuperPi 1M if you would like to compare them. But you shouldn't because these two benches are very different in its core. GPUPI has its own tweaks in hardware and software but like for every benchmark out there it depends on the bottlenecks to get the most efficieny.

 

But I get what you mean. GPUPI relies on another API+driver layer, like 3D benchmarks with DX and GPU driver. That adds another level of tweaking as well that should be used of course.

Edited by _mat_

  • Author

Sorry for the delay, just came back from my vacation. :)

 

Errors like these indicate a malformed file or an errornous upload. The submission files are very sensitive because of their encryption. As the browser switch solved your problem I would guess a browser extension got in the way.

WARNING - Full GPU noob.

 

So I've tried clocking on the OpenCL check box in GPU-Z and looked all through Catalyst Control Center for a box to check or turn on and can't find it.

 

Is it hiding somewhere else?

no need to search inside the control center.its in possibel to enabel opencl support via software if the hardware don't have it integrated.

 

ps: we al started some day as a noob

Nah nothing there, had another look.

 

Tried installing SDK separately, tried rolling back to 12.1 and 12.4 with older OpenCL in it.

 

Nothing.

 

A quick google search has revealed quite a few "miners" who had the same problem trying to mine with these cards. Maybe they just aren't OpenCL compatible?

 

Will keep searching around and see if there is a fix somewhere buried deep in the interwebs.

Got it to at least recognise the GPU's but still nothing.

 

Ar2rDLRh.png

 

Anyone come across this before and can steer me in the right direction?

  • Author

Although double precision was introduced with the HD 4800 series, it seems like it still does not support it via OpenCL. Maybe it's only a driver bug, but I don't think that the double precision extension was ever implemented for it. I remember these times, OpenCL was not around when the RV770 was launched in 2008 and also after it's first draft in August 2009 it took a while to catch on. Back then ATI/AMD had its own GPGPU languages like Brook and CTM (Close-To-Metal) ... both are no longer supported by the AMD Stream SDK.

 

So I guess you are out of luck with those graphics cards. :(

 

Btw, I have not given up on the idea to replace double precision with an float emulation mode yet. Let's see what the future brings.

skulstation -

what ar the chanses thad it whil work whit Pentium III / amd K6 and below ?

 

Next to none. IIRC the software driver need Core of Xeon CPU's. From release notes:

 

The OpenCL Runtime provides OpenCL CPU device support on the following processors:

• Intel® Core™ Processors

• Intel Xeon Processor E3, E5, E7 families

https://software.intel.com/file/450209/download

 

It still won't work at all for me, tough.

 

 

****

 

 

I wonder, what to do, when GPUPI just output the message "Could not start worker thread" ...

 

Could_not_start_worker_thread.jpg

 

Win7 lite, installed the latest OpenCL drivers for Intel (Core 2 Extreme X6800 utilized) and the Visual C++ Redistributable Packages for Visual Studio 2013 (x86), as mentioned there: https://www.overclockers.at/news/gpupi-international-support-thread

 

Any ideas?

 

I used Open CL runtime for CPU v15.1 x86 (Win7 x86 only) from Intel from there: https://software.intel.com/en-us/articles/opencl-drivers#win32

 

I give a try the x64 versions using normal x64 Win7 install. Cannot seems to find older that v15.1 Open CL x32 CPU driver. If anyone know, where to get it, I test it gladly.

 

...

 

Then I installed Win7 64bit, instaled the Microsoft Visual C++ libs 2013 x64, installed OpenCL 15.1 x64 and run GPUPI - again failure:

gpupi_x64_fail_again.jpg

 

Makes me wonder, if anyone else could run it (on CPU only, using the v15.1 OpenCL) and if I could run it, when my graphic card is recognized and have drivers. For some testing I use ATI Rage 128 PRO and there is not Win7 drivers for it, as far, as I searched. W98/98SE/ME - yes. Win2k - yes. WinXP - yes. Win7 - NO! Suxx. At least I made some Cinebench scores (could be better, if there is a driver for GFX card installed, on the 32bit version the score go from 8.13 to 8.17 just by installing GFX card drivers on Xeon X5650... :nana: )

 

Never the less, I did not managed it to work. Any ideas?

  • Author
HD 4800 should work just fine. Well, at least it did when using older GPUPI 2.0. http://hwbot.org/submission/2830026_havli_gpupi___1b_radeon_hd_4870_16min_54sec_963ms

 

Maybe I'll try it again tomorow with latest GPUPI version. :)

Then it's a driver issue, because the latest GPUPI has no changes in this regard.

 

skulstation, I will have a look at OpenCL initialization on these old CPUs. Maybe there is a compatibility issue with certain parameters when creating the context.

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...