This is very commonly talked about so its not just me and you likely already know about it.
during opencl activity (btc mining) when i had 1 card installed (6950) i would get 1% cpu usage, but now i have 2 installed i get 100% cpu usage on the core i set their affinity to.
i had the same problem with various driver and sdk versions not being able to fix it.
it creates alot of extra heat and power usage for apparently no good reason as the cpu shouldnt be used at all....
we really appreciate the flash video lockup fix in 11.7, that was very annoying.
a fix to this also would be the last widespread problem ive heard of...
Seriously guys! Do you give a damn about us?
Looks like AMD do not take us seriuosly .
When a smallest bug accur in one of games - they work very hard to release patches and hotfixes , because gamers ARE valued customers .
And I ? I've bought more AMD graphic cards than most of other people on planet , I know people who bought DOSENS of these cards for bitcoin mining purposes : we ARE a huge comunity with lot of buying power - why we are not treated as valued customers ?!
That's because bitcoin mining is dying.
yes, bitcoin mining is dying.
but, AMD did not take us seriuosly even at the time when bitcoin mining scene was healthy. i dont see a relation.
i am wondering how it is possible to organize and make so much noise (with advertisement) for an contest like the one on topcoder.com. i guess the participants on there just do not know how unstable AMDs OpenCL runtime or the Catalyst driver is.
Any updates on this bug?
This is a real shame. Not even one comment from AMD here if I'm not mistaken...
Perhaps AMD is giving up against CUDA, which is realy coming along. And now there is OpenAcc, too. Guess AMD + Linux isn't meant to be. If this bug isn't fixed within 2 weeks, AMD is dead for me for good.
Catalyst 11.11 for linux partially fixes the problem. CPU load is way lower now (~30% as compared to ~100% previously). GPU_USE_SYNC_OBJECTS is still a noop, thus this works regardless how the environment variable was set.
I almost regained the pre-2.5 speeds which is good. Well, CPU utilization is still much higher than before (~30% vs 5-6%) but it's much better now. This catalyst version is the best thing that came out of AMD for the last 6 months.
Another thing to mention (sorry for hijacking the thread) - somehow the IL compiler was improved and it does not generate excessive MOV instructions when working with long types. One of my kernels is doing SHA-512, where all operations are 64-bit, rather than 32-bit. Going from 11.10 to 11.11 increased the speed by more than 100%, which is very nice 🙂
Still I have that problem with the BFE_UINT optimization ((a&b)>>c). For some reason, there are now excessive MOVs here which have no valid explanation. With 2.4, this ended up as AND+SHR, now we have MOV+BFE, sometimes even two MOVs and a BFE. It's just slower. It hurts the performance of my DES kernels 😞
I have a long one and I am not very much willing to post it here. Is there some way to submit a bug report (kernel+ISA dump)?
Thanks again, I will collect what's needed and submit that tomorrow (it's 1:30 already here, need some sleep)
Any news with this bug, I have two 7870 cards and when using OpenCL for GPU my 2 CPU are 100% busy.
Any solutions ?
Nice work finding such a old thread. Let me try asking some people on the current status.
I tried few combinations:
Drivers 12.8, 13.1, 13.4, 13.6 and bitcoin miner (from GUI mining - updated Dec 3 with 7970 bugfix, also supports Stratum!) 20121203 / 20120219
All drivers above 12.8 give me 100% CPU usage after some time (never longer then 5min), also with 20121203 client after next few min/sec performance went down from Mhash to khash so decrease 1000 times....
Only combination of 12.8 and 20120209 gives me comfortable set to maining. Important is that 12.8 always gives max speed and when U run the some game doesn't slow down a game performance. All later drivers prefer
OpenCL engine rather then DirectX/OpenGL so simple U cannot play the game, U have to stop the 'miner'.
When I was checking that, I always uninstall old drivers by AMD uninstalling tool, after restart I used Driver Fusion, and manually deleted opencl.dll and amdocl.dll from Windows folders then after next restart I installed new drivers.
System: Win 7 Ultimate 64bit
CPU: AMD FX 8150
MOBO: ASUS M5A97 fresh BIOS
HDD: 3x512GB RAID 5
GFX: 2x 7870 1.1GHz and 1.0GHz (manufacturer overclocking)
Dowloaded your software on a multi-gpu setup, have a few questions though:
1. Is the source of bitcoin available? I did saw some kernel files, so i guess atleast kernels are available.
2. Is the 100% CPU usage issue happen with any kernel that run on a multi-gpu setup, or some specific kernel?
3. I tried to run the software on a multigpu set-up, but in the "device" textbox, i could select only one device(out of 1 CPU and 2 GPUs). How to run it for multi-gpu.
4. Also the software seems to rely on some external server, and it was not running for me (IMHO), with the default server. Getting Error: Object in JSON format could not read (or something similar). Any ideas?
4. I am planning to modify SimpleMultiDevice sample to reproduce the issue. Any comments? Maybe you wanna do that
Thursday, June 6, 2013, 7:53:16 AM, you wrote:
> 1. Is the source of bitcoin available? I did saw some kernel files,
> so i guess atleast kernels are available.
GUIMiner is only a frond end (GUI) for other miners, I use poclbm
which is writen in pyton, source there:
Also after installation OpenCL parts are in plain text in program DIR.
> 2. Is the 100% CPU usage issue happen with any kernel that run on a
> multi-gpu setup, or some specific kernel?
I use only poclbm, but in few posts lads had this problem as well.
> 3. I tried to run the software on a multigpu set-up, but in the
> "device" textbox, i could select only one device(out of 1 CPU and 2
> GPUs). How to run it for multi-gpu.
Go to File=>New Miner=>OpenCL, You will have a new tab for new 'Miner'.
> 4. Also the software seems to rely on some external server, and it
> was not running for me (IMHO), with the default server. Getting
> Error: Object in JSON format could not read (or something similar). Any ideas?
U have to register, btcguild.com is OK.
If U don't want use my:
anglik666_4 and anglik666_5
pass is any (must be some but is whichever).
Server: Any BTCGuild from the list.
> 4. I am planning to modify SimpleMultiDevice sample to reproduce
> the issue. Any comments? Maybe you wanna do that
Sounds to me like a Mars expedition, maybe better I'll go for a pint
and U do that ?
> Reply to this message by replying to this email -or- go to the message on AMD Developer Forums
> SHARE THE LOVE! Help other developers by marking replies to your
> questions as Helpful or Correct. If you are the original question
> asker, you can mark replies as Helpful or Correct by:
> Go to the message on AMD Developer Forums (you’ll need to log in).
> Click on either the Helpful Answer button or the Correct Answer
> button. Pat yourself on the back! You’ve helped others who have the
> same question as you by telling them which replies are helpful and correct.
This is cool. I did not know you could just reply to the e-mail to post here....
but yeah -- It looks like it gives away your e-mail address though...
Hi, I also have an issue with 100% CPU load
Bug 1099 – One AMD CPU core is always at 100% load while Chrome is just displaying a blank page