Einstein@Home GPU/APU Application for AMD/ATI Graphics Cards: discussion thread

europa
europa
Joined: 29 Oct 10
Posts: 49
Credit: 34029952
RAC: 0

RE: RE: I just scrolled

Quote:
Quote:

I just scrolled through this thread but things are pretty dated so I'm hoping I can get some current help.

I'm converting a machine over to BOINC. It has an ATI HD7850 and a HD 7950 card and I'm running Linux Mint 17 (based on Ubuntu).

At this point, BOINC cpu WU's are running but it doesn't see the GPU's. I made the change to cc_config.xml. I have NOT installed any proprietary drivers. I looked around for Catalyst 12 since they were mentioned earlier but can't find it anymore although I do have Catalyst 13 and 14 in my archives along with the SDK.

Should I install one of these or can I get what I need from the repository? What would those files be? Also, what about installing the SDK?

Thanks for any help you can provide. All my experience is with Nvidia.

Regards,
Steve

Looks like your AMD GPU has been identified and that you are up and running. Congrats

Robi/Gary,

Thanks for your suggestions but unfortunately things haven't worked. Although I can get things working immediately after I install BOINC (only after installing the ATI drivers first), as soon as re-boot, it no longer sees the GPUs.

It looks like the problem lies, at least in part, with the coproc_info.xml file but even copying over a good copy doesn't solve the problem.

I've run out of ideas and it's no longer worth the aggravation so I'll stick with my other Nvidia machines.

Anyway, thanks again for your ideas.

Regards,
Steve

Claggy
Claggy
Joined: 29 Dec 06
Posts: 560
Credit: 2694028
RAC: 0

RE: Thanks for your

Quote:
Thanks for your suggestions but unfortunately things haven't worked. Although I can get things working immediately after I install BOINC (only after installing the ATI drivers first), as soon as re-boot, it no longer sees the GPUs.


How about:

Debian/Ubuntu/Mint/Derivatives - GPU recognition fixes

Quote:

Issues & Solutions:

ISSUE #1: BOINC starts before GDM can finish getting up and running. This makes the BOINC think the video card is absent. This can occur with both ATI and Nvidia cards but I first found it on Nvidia card equipped machines. If your GPU is recognized after you restart BOINC but doesn't find the card right after a reboot this is likely your problem. This can be a bit sporadic on some machines where it'll find it after one reboot but not another. If this is the only problem it'll find it every time after a restart of just BOINC.
sudo /etc/init.d/boinc-client restart

See history here: https://bugs.launchpad.net/ubuntu/+source/boinc/+bug/414244

FIX #1: My fix was to add a delay into the start-up script so that GDM/X has a chance to have drivers fully loaded. This solution inserts a 6 second delay in the boinc start-up script which allows ample time for the driver to be available on all machines tested to date.
(a)Edit the start-up script file /etc/init.d/boinc-client with sudo gedit /etc/init.d/boinc-client.
(b)Find this function code and add the line “sleep 6†where it's shown here:
start() { log_begin_msg "Starting $DESC: $NAME" if is_running; then log_progress_msg "already running" else sleep 6 start-stop-daemon --start --quiet --background --pidfile $PIDFILE \ --make-pidfile --user $BOINC_USER --chuid $BOINC_USER \ --chdir $BOINC_DIR --exec $BOINC_CLIENT -- $BOINC_OPTS fi log_end_msg 0 if [ "$SCHEDULE" = "1" ]; then schedule fi }

FIX #2: Modify the start-up sequence of the /etc/init.d/boinc-client script.
(a)I got this from gfarmerfr an ATI/Ubuntu user on the DNETC forums after telling him how I was putting the delay in. I haven't actually tried it. It worked for gfarmerfr and he knows what he's doing. I believe this was tested on v9.10. In many ways this is a better fix but I was already into the script for other reasons (adding fan speed control) so I stuck with my sleep 6 version.
(b)Execute the following two commands to move BOINC start-up to the end of start-up processing:
sudo update-rc.d -f boinc-client remove sudo update-rc.d boinc-client defaults 99

Claggy

robl
robl
Joined: 2 Jan 13
Posts: 1709
Credit: 1454573908
RAC: 3202

RE: RE: RE: I just

Quote:
Quote:
Quote:

I just scrolled through this thread but things are pretty dated so I'm hoping I can get some current help.

I'm converting a machine over to BOINC. It has an ATI HD7850 and a HD 7950 card and I'm running Linux Mint 17 (based on Ubuntu).

At this point, BOINC cpu WU's are running but it doesn't see the GPU's. I made the change to cc_config.xml. I have NOT installed any proprietary drivers. I looked around for Catalyst 12 since they were mentioned earlier but can't find it anymore although I do have Catalyst 13 and 14 in my archives along with the SDK.

Should I install one of these or can I get what I need from the repository? What would those files be? Also, what about installing the SDK?

Thanks for any help you can provide. All my experience is with Nvidia.

Regards,
Steve

Looks like your AMD GPU has been identified and that you are up and running. Congrats

Robi/Gary,

Thanks for your suggestions but unfortunately things haven't worked. Although I can get things working immediately after I install BOINC (only after installing the ATI drivers first), as soon as re-boot, it no longer sees the GPUs.

It looks like the problem lies, at least in part, with the coproc_info.xml file but even copying over a good copy doesn't solve the problem.

I've run out of ideas and it's no longer worth the aggravation so I'll stick with my other Nvidia machines.

Anyway, thanks again for your ideas.

Regards,
Steve

Before you give up.

I am assuming you are using Mint's BOINC. If this is true remove it and download and install BOINC from Berkely. This will however require that you manually start it after a reboot unless you are familiar with setting up a file to auto start it. My machines only need a reboot when a new kernel is installed so a manual restart for me is not a big deal.

I noticed that you machine correctly IDs the AMD GPU but there are no credits for the last few days. Are you processing GPU WUs? Do you have the AMD checkbox selected in your profile at the E@H site?

I am seeing this entry in your computer logs:

2015-01-21 13:59:59.5475 [PID=10665]2015-01-21 13:59:59.2571 [PID=10662] SCHEDULER_REQUEST::parse(): unrecognized: 0

Not sure if this might be contributory.

If a manual restart after reboot is not to your liking then ignore what I have posted above. Claggy's post might be helpful. I myself have not encountered these problems.

europa
europa
Joined: 29 Oct 10
Posts: 49
Credit: 34029952
RAC: 0

RE: RE: Thanks for your

Quote:
Quote:
Thanks for your suggestions but unfortunately things haven't worked. Although I can get things working immediately after I install BOINC (only after installing the ATI drivers first), as soon as re-boot, it no longer sees the GPUs.

How about:

Debian/Ubuntu/Mint/Derivatives - GPU recognition fixes

Quote:

Issues & Solutions:

ISSUE #1: BOINC starts before GDM can finish getting up and running. This makes the BOINC think the video card is absent. This can occur with both ATI and Nvidia cards but I first found it on Nvidia card equipped machines. If your GPU is recognized after you restart BOINC but doesn't find the card right after a reboot this is likely your problem. This can be a bit sporadic on some machines where it'll find it after one reboot but not another. If this is the only problem it'll find it every time after a restart of just BOINC.
sudo /etc/init.d/boinc-client restart

See history here: https://bugs.launchpad.net/ubuntu/+source/boinc/+bug/414244

FIX #1: My fix was to add a delay into the start-up script so that GDM/X has a chance to have drivers fully loaded. This solution inserts a 6 second delay in the boinc start-up script which allows ample time for the driver to be available on all machines tested to date.
(a)Edit the start-up script file /etc/init.d/boinc-client with sudo gedit /etc/init.d/boinc-client.
(b)Find this function code and add the line “sleep 6†where it's shown here:
start() { log_begin_msg "Starting $DESC: $NAME" if is_running; then log_progress_msg "already running" else sleep 6 start-stop-daemon --start --quiet --background --pidfile $PIDFILE \ --make-pidfile --user $BOINC_USER --chuid $BOINC_USER \ --chdir $BOINC_DIR --exec $BOINC_CLIENT -- $BOINC_OPTS fi log_end_msg 0 if [ "$SCHEDULE" = "1" ]; then schedule fi }

FIX #2: Modify the start-up sequence of the /etc/init.d/boinc-client script.
(a)I got this from gfarmerfr an ATI/Ubuntu user on the DNETC forums after telling him how I was putting the delay in. I haven't actually tried it. It worked for gfarmerfr and he knows what he's doing. I believe this was tested on v9.10. In many ways this is a better fix but I was already into the script for other reasons (adding fan speed control) so I stuck with my sleep 6 version.
(b)Execute the following two commands to move BOINC start-up to the end of start-up processing:
sudo update-rc.d -f boinc-client remove sudo update-rc.d boinc-client defaults 99

Claggy

Claggy,

Thanks for the suggestions but still no luck. I remember that problem well from many years ago but haven't had it since then with the Nvidias through multiple new installs, etc. Given how many people had the problem, I would think that they fixed it.

I did notice that with your script change for adding 6 sec., the boinc-client init.d file appeared to already call for 10 sec. delay. I'm not a coder so I may not be reading the line correctly. Also, the second solution appeared to delete a number of symbolic links.

I'll let it sit for a bit and see if there's anything else that I can try.

Thanks again for the input.

Regards,
Steve

europa
europa
Joined: 29 Oct 10
Posts: 49
Credit: 34029952
RAC: 0

RE: RE: RE: RE: I

Quote:
Quote:
Quote:
Quote:

I just scrolled through this thread but things are pretty dated so I'm hoping I can get some current help.

I'm converting a machine over to BOINC. It has an ATI HD7850 and a HD 7950 card and I'm running Linux Mint 17 (based on Ubuntu).

At this point, BOINC cpu WU's are running but it doesn't see the GPU's. I made the change to cc_config.xml. I have NOT installed any proprietary drivers. I looked around for Catalyst 12 since they were mentioned earlier but can't find it anymore although I do have Catalyst 13 and 14 in my archives along with the SDK.

Should I install one of these or can I get what I need from the repository? What would those files be? Also, what about installing the SDK?

Thanks for any help you can provide. All my experience is with Nvidia.

Regards,
Steve

Looks like your AMD GPU has been identified and that you are up and running. Congrats

Robi/Gary,

Thanks for your suggestions but unfortunately things haven't worked. Although I can get things working immediately after I install BOINC (only after installing the ATI drivers first), as soon as re-boot, it no longer sees the GPUs.

It looks like the problem lies, at least in part, with the coproc_info.xml file but even copying over a good copy doesn't solve the problem.

I've run out of ideas and it's no longer worth the aggravation so I'll stick with my other Nvidia machines.

Anyway, thanks again for your ideas.

Regards,
Steve

Before you give up.

I am assuming you are using Mint's BOINC. If this is true remove it and download and install BOINC from Berkely. This will however require that you manually start it after a reboot unless you are familiar with setting up a file to auto start it. My machines only need a reboot when a new kernel is installed so a manual restart for me is not a big deal.

I noticed that you machine correctly IDs the AMD GPU but there are no credits for the last few days. Are you processing GPU WUs? Do you have the AMD checkbox selected in your profile at the E@H site?

I am seeing this entry in your computer logs:

2015-01-21 13:59:59.5475 [PID=10665]2015-01-21 13:59:59.2571 [PID=10662] SCHEDULER_REQUEST::parse(): unrecognized: 0

Not sure if this might be contributory.

If a manual restart after reboot is not to your liking then ignore what I have posted above. Claggy's post might be helpful. I myself have not encountered these problems.

Robi,

That seems to have done it. I've re-booted a couple of times and am processing WU's. E@H is sending GPU WU's although the others haven't so far. My next stop to to check my preferences at MW@H and Seti.

One thing I am noticing with E@H, is that the progress keeps going back and forth between normal processing and suspended CPU busy. I'm guessing that I have to adjust the ratio of CPU processing to GPU (e.g., .75cpu to, say, .25gpu) to account for the greater GPU memory on the ATI cards vs. the Nvidia's.

Thanks again for the help.

Regards,
Steve

europa
europa
Joined: 29 Oct 10
Posts: 49
Credit: 34029952
RAC: 0

One follow-up question. I

One follow-up question. I notice that E@H is displaying the notation (0.5cpu and 0.25ati Etc.) after the gpu units. I know how to adjust the GPU Wu's but can't find where to change the CPU fraction to increase it since E@H no longer uses app_info.xml files.

Can someone point me in the right direction?

Thanks for the help.

Regards,
Steve

Richard Haselgrove
Richard Haselgrove
Joined: 10 Dec 05
Posts: 2142
Credit: 2775640871
RAC: 811660

RE: One follow-up question.

Quote:

One follow-up question. I notice that E@H is displaying the notation (0.5cpu and 0.25ati Etc.) after the gpu units. I know how to adjust the GPU Wu's but can't find where to change the CPU fraction to increase it since E@H no longer uses app_info.xml files.

Can someone point me in the right direction?


Projects never use app_info.xml files - they exist for users with non-standard applications or platforms to create if they wish. I wouldn't recommend it, though, unless you have a pressing need.

You can change the declared CPU usage if you wish with an app_config.xml file (see Application configuration), but I'm not sure that's what you really want. It only affects how many CPU tasks are scheduled to run alongside your AMD/ATI task - it doesn't directly affect the running of the AMD/ATI task itself.

In your previous post, you mentioned "going back and forth between normal processing and "suspended CPU busy'". That suggests that you have your global computing preferences set too low: visit Computing preferences, and verify

Quote:
Suspend work if CPU usage is above
0 means no restriction
Enforced by version 6.10.30+


(or if you are one of the people who prefers over-riding the web preferences via local settings set by BOINC Manager, check the local version of the 'processor usage' figure instead)

europa
europa
Joined: 29 Oct 10
Posts: 49
Credit: 34029952
RAC: 0

RE: RE: One follow-up

Quote:
Quote:

One follow-up question. I notice that E@H is displaying the notation (0.5cpu and 0.25ati Etc.) after the gpu units. I know how to adjust the GPU Wu's but can't find where to change the CPU fraction to increase it since E@H no longer uses app_info.xml files.

Can someone point me in the right direction?


Projects never use app_info.xml files - they exist for users with non-standard applications or platforms to create if they wish. I wouldn't recommend it, though, unless you have a pressing need.

You can change the declared CPU usage if you wish with an app_config.xml file (see Application configuration), but I'm not sure that's what you really want. It only affects how many CPU tasks are scheduled to run alongside your AMD/ATI task - it doesn't directly affect the running of the AMD/ATI task itself.

In your previous post, you mentioned "going back and forth between normal processing and "suspended CPU busy'". That suggests that you have your global computing preferences set too low: visit Computing preferences, and verify

Quote:
Suspend work if CPU usage is above
0 means no restriction
Enforced by version 6.10.30+

(or if you are one of the people who prefers over-riding the web preferences via local settings set by BOINC Manager, check the local version of the 'processor usage' figure instead)

Richard,

Thank you. That seems to have done it. I changed both entries to "0."

The notation still remains after each GPU WU about 0.5 cpu etc. but I can't figure out where it's coming from.

The WU's appear to be cranking steadily now, so I'm happy.

Regards,
Steve

robl
robl
Joined: 2 Jan 13
Posts: 1709
Credit: 1454573908
RAC: 3202

RE: Before you give up.

Quote:

Before you give up.

I am assuming you are using Mint's BOINC. If this is true remove it and download and install BOINC from Berkely. This will however require that you manually start it after a reboot unless you are familiar with setting up a file to auto start it. My machines only need a reboot when a new kernel is installed so a manual restart for me is not a big deal.

I noticed that you machine correctly IDs the AMD GPU but there are no credits for the last few days. Are you processing GPU WUs? Do you have the AMD checkbox selected in your profile at the E@H site?

I am seeing this entry in your computer logs:

2015-01-21 13:59:59.5475 [PID=10665]2015-01-21 13:59:59.2571 [PID=10662] SCHEDULER_REQUEST::parse(): unrecognized: 0

Not sure if this might be contributory.

If a manual restart after reboot is not to your liking then ignore what I have posted above. Claggy's post might be helpful. I myself have not encountered these problems.

Robi,

That seems to have done it. I've re-booted a couple of times and am processing WU's. E@H is sending GPU WU's although the others haven't so far. My next stop to to check my preferences at MW@H and Seti.

One thing I am noticing with E@H, is that the progress keeps going back and forth between normal processing and suspended CPU busy. I'm guessing that I have to adjust the ratio of CPU processing to GPU (e.g., .75cpu to, say, .25gpu) to account for the greater GPU memory on the ATI cards vs. the Nvidia's.

Thanks again for the help.

Regards,
Steve

You need to reserve a cpu to feed the GPU. In the Boinc manager under tools->preferences at the bottom is a "multi processor use some%" try 60 percent to see if this fixes your "suspended" issue. If it does try increasing to 80%. You might see a cpu WU go into a "wait" state. That is ok. Just leave it alone and BOINC will handle it.

I am set at 80% which crunches 4 cpu WUs and 4 GPU WUs. Wuuuu! Wuuu!

europa
europa
Joined: 29 Oct 10
Posts: 49
Credit: 34029952
RAC: 0

Robi, Thanks for the

Robi,

Thanks for the advice. I'll do it as soon as I fix the machine. A little bit ago there was suddenly a sharp bang and now it's not running. It's always something!

Regards,
Steve

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.