Hello!
some years i use my nvidia GT630 videocard with nvidia drivers to crunch for seti via boinc.

i use gpu tasks only.
when i start that, i also see high cpu usage, but find i may use libsleep, and since then my cpu usage is about at 2-3% level at one core only, for seti.

all is good till i upgrade that machine to last stable slackware version - slackware64 14.2, kernel 4.4.38 .
since this, i found, my cpu again run on 100% (one core) all the time, yet that all was gpu tasks.

looks like libsleep not work anymore.
there is any workarounds? not want to run cpu on high load for nothing ( all work do GPU anyway)....electricity not cheap.

---

1) what PC you are talking of? link. I see only 2 OS X based PCs under your account.
2) -use_sleep if it's OpenCL GPU app.

---

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

https://setiathome.berkeley.edu/hosts_user.php?userid=1057334

this one.

2. may help me in this? try to google more about it, thanks...

---

BOINC# ./boinc -use_sleep
./boinc: /lib64/libssl.so.1.0.0: no version information available (required by ./boinc)
./boinc: /usr/lib64/libcurl.so.4: no version information available (required by ./boinc)
./boinc: /lib64/libcrypto.so.1.0.0: no version information available (required by ./boinc)
Unknown option: -use_sleep

---

that's app option, not boinc option. Look ReadMe files in project directory.

---

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

that means, manual run each task?

nit fit in my case, anyway. ill see this month, what is electricity bill, and if it considerably higher, than in last months, i simple stop crunching on this machine.
earlier libsleep make a good work, but looks like it not work anymore, and no one knows why, and what restore low cpu load on gpu tasks.....
sadly, but...

---

http://setiathome.berkeley.edu/forum_thread.php?id=74160

old thread about that theme.

and from /boinc/%projectdir%astropulse_7.08_README_x86_64-pc-linux-gnu__opencl_nvidia_100


-use_sleep :Results in additional Sleep() calls to yield CPU to other processes. Can affect performance. Set by default on 64bit Linux for nVidia GPUs.^M

it says -use_sleep option is used by default, as so, there seems no sense in use additional -use_sleep switch in linux64 OS case...?

---

BOINC# ./boinc -use_sleep
./boinc: /lib64/libssl.so.1.0.0: no version information available (required by ./boinc)
./boinc: /usr/lib64/libcurl.so.4: no version information available (required by ./boinc)
./boinc: /lib64/libcrypto.so.1.0.0: no version information available (required by ./boinc)
Unknown option: -use_sleep

The version of BOINC you're using doesn't work very well with the newer versions of Linux, is there any way you can update BOINC?
If that doesn't work, you could just use the older version of CUDA under Anonymous platform. The CUDA 4.2 App uses little CPU and would probably work just as well as the newer Apps on that low end GPU.
You can find the package here; http://www.arkayn.us/forum/index.php?topic=197.msg4489#msg4489
If you don't want to use the CPU, remove the CPU section from the bottom of the app_info.xml file.

that means, manual run each task?

nit fit in my case, anyway. ill see this month, what is electricity bill, and if it considerably higher, than in last months, i simple stop crunching on this machine.
earlier libsleep make a good work, but looks like it not work anymore, and no one knows why, and what restore low cpu load on gpu tasks.....
sadly, but...

it does mean to add option for app, BOINC will launch app as it did before.
But for Linux -use_sleep is default option indeed (worth to refer MultiBeam ReadMe instead of AstroPulse though).
That could mean smth should be done with OS itself for process to correctly yield CPU on request.
It's the first report of such issue so some experimenting would be needed.
For example, to go to anonymous platform and install CUDA app insteaad of OpenCL one as TBar suggested. Of course it's for MultiBeam, not AstroPulse, but AstroPulse so scarce that can be neglected for now.

Will be interesting to see if it experience any issues with CPU yielding (though driver-level yields can't be equal process-level ones it could give some additional info).

---

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

Hello!
some years i use my nvidia GT630 videocard with nvidia drivers to crunch for seti via boinc.

i use gpu tasks only.
when i start that, i also see high cpu usage, but find i may use libsleep, and since then my cpu usage is about at 2-3% level at one core only, for seti.

all is good till i upgrade that machine to last stable slackware version - slackware64 14.2, kernel 4.4.38 .
since this, i found, my cpu again run on 100% (one core) all the time, yet that all was gpu tasks.

looks like libsleep not work anymore.
there is any workarounds? not want to run cpu on high load for nothing ( all work do GPU anyway)....electricity not cheap.

Hi,
I do not see any NVIDIA cards listed on any of your computers... You may need to load a driver for the GTX630 first.

---

To overcome Heisenbergs:
"You can't always get what you want / but if you try sometimes you just might find / you get what you need." -- Rolling Stones

Hi,
I do not see any NVIDIA cards listed on any of your computers... You may need to load a driver for the GTX630 first.

It's under a different account, https://setiathome.berkeley.edu/hosts_user.php?userid=1057334. ;-)

Cheers.