Re: GPU Temp. for 2 video cards
Bill WIlson <[email protected]>
| Newsgroups | gmane.comp.gnome.apps.gkrellm |
|---|---|
| Message-ID | <[email protected]> |
On Wed, 28 Mar 2007 18:19:59 -0500 Bill WIlson <[email protected]> wrote: > I've run into a problem with this. nvclock eats a big chunk of CPU and it > adds 4 seconds to the gkrellm start up time. I found that nvclock has a problem with at least the nvidia chipset (Geforce 6150) on a couple of my boxes. It reads the video card bios and stores a copy in ~/.nvclock/bios0.rom and subsequent runs use this cached bios as long as it passes some checks to verify it's still valid for the found video card. nvclock with my 6150 always fails in this check and so the bios is reread every time nvclock is run and this causes my high CPU usage. So, I can't do an automatic nvclock check and there will need to be some kind of configuration enabling its use. But in the meantime, I've gone ahead and put up a tarball you can use for nvclock gpu temps that does do an automatic check when nvidia-settings fails to find any temps. I'm still thinking this over, but I expect the way it will work is when you first upgrade to any gkrellm version >= 2.2.11, you'll have to go into the sensor config somewhere and enable using nvclock. The tarball is: http://members.dslextreme.com/users/billw/test/gkrellm.tar.gz Let me know of any problems if you try it. Bill _______________________________________________ Gkrellm mailing list [email protected] http://lists.jutley.org/cgi-bin/mailman/listinfo/gkrellm