Some one can explain me why this funder 7900XTX crash on tons of my games please ? and why im loosing so many perf ?
Lost ark = crash (70% utilization and power/loose frames)
Warzone 2 = crash (100% utilization)
Naraka = crash (100% utilization)
Warhammer draktide = work fine (100% utilization)
conqueror blade = work "fine" (70% utilization and power/loose frames)
Cyberpunk = seems to work "well" but I had 1 crash yesterday (100% utilization)
Deep rock galactic = work "fine" (70% utilization and power/loose frames)
No crash in 3D Mark benchmark
No crash in occt cpu/GPU stress test (before junction temps go to up and stop the test...)
50% of time when one game crash i need to restart my computer because lot of softwares not working as intended... some visual bug happen in my windows to. Using DDU before last driver in safe mode.
My spec :
Windows 11 22H2
ROG strix Z690-A gaming wifi d4
32GO ddr4 3200mhz
Corsair RM1000x (1000w power)
Yes my CG have a 110c junction temp.
All work great before i use this CG with my RTX 2060.
Same issue here.
Similar specs except memory and mb. (2x16gb 3600cl16 ddr4, ASUS Prime Z690M-PLUS D4)
Crashes in pretty much every game where fully utilized. (Spider Man Remastered, The Witcher 3 Next-Gen, Battlefield 2042, RDR2)
Junction temp max is around 85c tho.
Same issue for me too, i thought it was limited to FFXIV, but i had crashes in Demeo, Neon white, Ghostwire Tokyo...
Did you remove using DDU, the old drivers?
If you have a junction temp of 110oC and the difference in temperature between junction temp and GPU core temp is >25oC you have a faulty cooler (assuming you have a reference card) - You should RMA your card IMEDIATELY.
i would concentrate on 110°C junction , check if fans a 100% , if fans are 100% and you have good air flow and junctions temps still 110° thats problematic and cause throttling ...
but 25°C difference is nothing , mine works perfectly fine with >30 ° difference, you can be far from 110 with a difference of 35
There is definitely something wrong with these cards. I bought a reference card with 110 junction temps and games crashing. RMA returned the card and went back to my 1070 with zero crashing. While in the return process I managed to snag an XFX Merc 7900 XTX AIB card thinking my reference was the issue. As soon as I installed it I'm back to constant crashing on basically any game I play.
At this point I'm assuming it's driver related?
you give to small info ... but maybe with the merc you did not a proper installation for example ... "reset bios defaults" after changing gpu board , and "reset factory default" or "DDU" in windows for driver installation, that could explain a lot and can cause problems when changing from one gpu brand to another
my asus 7900xtx MBA works perfectly so , no there are maybe some that have a 110 junction temps problem, but clearly a limited amount for what it seems .. and trolls saying it's a "design" problem don't seem to understand that if it was a design problem every card would be defective not just some of them , i know several people not having any issue with the mba , and merc aio is clearly not affected , just see notation of this card on some well known seller website i think it's almost 5 stars... so ...
for the rest there are "known issues" investigated by amd , you can find a list in the driver info to see if maybe you are concerned by these and maybe find a temporary solution
ASUS TUF X570 Plus, AMD 5900x, XFX Merc310 7900 XTX, 64 GB of Ripjaws 3600 DDR4 RAM, Superflower 1200 Watt platinum PSU, Windows 10 Pro playing on a Samsung EVO 1TB 980 NVME drive. Freshly reformatted windows, I always use DDU, not that it matters on a reformatted drive. God of War 2018 is fine for usually 40 minutes or so and then will hang and crash to desktop at random times beyond that time. Once it crashes the first time if I reload the game without restarting my PC first I will get crashes randomly after 25 minutes or so. BIOS is up to date, Drivers are now 23.1.1 (which didn't fix it from the prior December drivers).
high refresh rate monitor ?
just something for you's to try is to turn off XMP.
my friend had the same issues with there RX 7900 XTX and we went through all solutions and could seem to find the problem but with no luck. we then tried to see if turning off XMP might help (i know a random choice) but it actually fixed the issue so i thought to see if this helps you's. just to clarify this helped with games that did and didn't suffer driver crashes.
good clue, your friend should check if his bios is up to date , then if still not ok, try to set ram timings manually
yeah it worked for us so might work for him
we also tested different refresh rates thinking that was the issue but found changing XMP was the only solution so far that worked.
setting timings manually is entering values from the stick of rams, not just modifiying refresh rate
cas latency - RAS to CAS - RAS Precharge - tRAS ... etc
if you don't know about it, simpliest thing is to ask someone to do it for you or to wait for a bios update that improve stability (maybe check if ram sticks are in QVL list for your motherboard)
I've been getting about 1 crash an hour playing FF14. It's the only one that causes the driver itself to crash and reset. It gives a fatal DX11 error.
Same for me all the far cry games are not working sent ne card back to over clocker thay tested it sead it was not faulty. this my settup. if put my old card back in it works fine that RX580
ASUS X470 ROG Crosshair VII Hero
AMD Ryzen 7 2700X
Team Group Vulcan T-Force 16GB (2x8GB) DDR4 PC4-
Kolink KL-1000M 1000W 80 Plus Bronze Modular Power Supply
Sapphire Radeon RX 7900 XTX Nitro+ Vapor-X 24GB GDDR6 PCI-Express Graphics Card
and latest bios on 5003
Overclock is blocking 100%
I had the same problems, I downgraded my gpu driver to 23.1.1 and everything is fixed if the games you play have Upscale rendering option use Intel XESS or turn it off.
Looks like the Kolink KL-1000M are **bleep** had get good one. gone for a 1300w Antec and it seems to be ok,
Man, and i adore my 7900xtx ref card. Also, even if your hitting 110, you should not crash, it would just throttle. Then, you have 2 cards crash,... imho, its not the cards, you have something else going on. 119 temps, still don't cause crashes. Just throttles. That, is out of the equation , as well you saying you tried the other card, and didn't bring up temps. Temps don't cause crashes your describing. The card would run all day at 110, it would throttle though. You need to look else where imho. Even changing to your old 2060. If i was you, id look more at your PSU.