• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

53.28 drivers (Linux) o/c unfriendly?

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

Texas Gore

Registered
Joined
Dec 15, 2003
Location
not entirely sure...
I recently installed the 53.28 drivers on my Gentoo box and found they are unstable at my current GPU/RAM clock speeds on GF4 Ti4200 (320/620).

I installed the driver (available from the nvidia website ) and decided to run a couple of benchies (glxgears and Q3A Timedemo) to see if performance had increased. The first sign of instability came when quitting glxgears...hard system lock-up, requiring ye olde reset button :(

The next sign came with agonisingly loooooong pauses in Q3A Timedemo (usually a symptom of GPU clock too high in my experience) and also visual artefacts... Lastly, a hard system lock-up when quitting Q3A, again requiring ye olde reset...

My first thoughts of "buggy driver" went, and thoughts of "driver doesnt like my o/c" came... confirmed when I reduced clock speeds by 10MHz GPU and 20MHz RAM (310/600) - artefacts remained when RAM reduced by 10MHz.

Has this happened to anybody else? Is there a similar occurance in the win32 driver? Or is it something in my system...

Im curious...
 
I have seen this problem in the past with both Linux and windows drivers. It seems that an increase in driver efficiency through better optimization keeps the GPU pipelines fuller ( busier) and that means more current drawn and higher temps.

Your card probably starts failing at around the same temp as before, but is reaching that temp at a lower clock speed.
 
Thanks for the reply.

Higher optimisation and busier pipes make sense, so that may explain it.

However, the performance gained by the driver "optimisations" falls way below that of the increased clocks on older drivers (30-35 fps Q3 timedemo), so I'll stick with the 4496 for now. ;)
 
Back