VR ctds pink lines in vrfps bam^ crash.

No matter what l do. It crashes to desktop or just freezes.
I've done undervolting, thermals, fans,
Ambient temp now 47°.
Gpu temp under load about 50 - 73%
Elite odyssey works perfectly in pancake. Although I hate it. Tests perfectly.
I can play sometimes mission running, for 15 minutes or more, then vrfps which at that point (approaching a station or jumping systems etc) starts spiking pink lines rapidly....the all sweet green lines go solid red with pink lines, bam! Freezes.
Temps about 60° or so.
Steam slider at 50% then raised to 100% in /steamvr/video/per application video settings.
Ss at .65
Hmd 2.0
Vr medium selected with a few minor changes but the big ones like bloom blur fx quality ambient occlusion etc are all off. Fx quality I'll push to medium if l can.
I've a 3070ti 9900k 32gig rig with a g2 reverb v1.
Any help would be appreciated.
 
Last edited:
Ampere GPUs generate a lot of heat, have you good airflow in your case?

For reference with my 3080 I use 3 x 140mm front case fans in a Phanteks and I found I had to up my fan rpm curve otherwise I got CTD and freezes
 
ive 3x140mm fans (fractal designs came with the case) at the front sucking air in.
ive a 140mm fan at the base (with a nice front cover over it so it dont chew cables above it) sucking air in. 850w seasonic psu is mounted bottom rear sucking air in obviously. ive two 140mm fans drawing air out of the top and top rear. finally the cpu aio at the top (its ajust a 140mm aio).
ambient temp idling is about 45 -50 degrees C. cpu idles at 28 to 30 degrees. gpu about 50 to 52.
stressed at stock about 73 degrees max. ive upped the power to 105% so its about 75 now. heaven bench and superpostition stress n bench never get it above 75.
id love to config my own curve in afterburner ive tried it dont work....doesnt do a thing even the firmware option does nowt.
thinking of using smartfan in BIOS?
 
Temperatures got nothing to do with that. The only part in a system that could suffer from temp-induced instability might be the PSU if it can't mantain a stable load, leading to cascade instability to other components. An overheating CPU or GPU "usually" will just throttle down, or reach thermal shutdown, but they need to overheat pretty badly to reach that point (we are talking temps around or above 100 °C). First thing I'd suggest is to start back from a "safe" profile, no voltage/frequency tweaks for any of the components, stock power profile, even disabling XMP if enabled, just to be on the safe side, and see what happens from there. No point in chasing a cooling mirage if a machine is acting up at just 75 °C on CPU/GPU.
 
ok ive enabled a smartfan manual setup whereby they come on much more vigorously and earlier too. sounds ok though.
just run cinebenchs heaven 4.0 and superposition in vr res at same time. didnt go over 74 degrees
 
The first thing you need to do is determine if it’s hardware or software, try different VR games.
The Lab from Steam is free and a lot of fun.
 
wow those fans in manual made a bit of a racket under extreme load.
so switched to auto in smartfan. much better.
so went ingame as per settings in Op and yeah seems ok. got a skipload of pink n red at bases as expected. but didnt crash!
reprojection virtually 0 till bases then about 25% drops wen i leave or jump.
ive flown in VTOL = perfect nps wotsoever...also flatscreen elite nps wotsoever
 
From my experience, the new WMR for Stream update introduced a new VRAM usage optimization may involved. It may cause VRAM overheat time to time in some games. You can use WMR for Stream beta version to enable the option for disable it.
 
Back
Top Bottom