Random Shutdowns with R9 290X Reference Card

emperor886

Reputable
Jan 5, 2015
3
0
4,510
Hi there,



I'm coming here for help with random shutdowns using my R9 290X. It is the reference card from AMD, came in my Alienware Aurora R4 (actually, they replaced it with a new one just three days ago, as well as the power supply and motherboard). I have had intermittent issues since day 1, and honestly I think I know the cause of the issue, but I'm coming here for a second opinion seeing as Alienware tech support tells me I'm wrong.



So, first off:

Speccy: http://pastebin.com/8x51gmse

DxDiag: http://pastebin.com/Aq5yNymx

AlienAutopsy Log: http://pastebin.com/qfAtD062

Power Supply: http://imgur.com/KHP42wt

Others: http://imgur.com/KXxsiyb



I'm pretty sure the issue is simply that the PSU isn't providing enough Amps on the 12v rail. But..a second opinion would be great, if anyone has the time, please and thanks.



Oh, guess I should make clear, the shutdowns are random during gaming/benchmarking (Can't even complete 3DMark Ice Storm, the benchmark for smartphones...). Oddly enough, Folding@Home can run for, well, ever, with no problems. Shutdowns occur whether I'm at stock voltage/clocks or OC settings.

Edit: Just wanted to add, it may be the reference card but I have no issues with cooling whatsoever. Stock clocks, max temp 62, OC max temp is 70 (GPU + all system fans 100%, idc about noise, actually sounds kinda of cool like an airliner in my bedroom lol).
 
Have you tried upping your voltage, or changing which power supply leads the card is attached to? Might be simple as that. I agree though, sounds like you're just not getting enough juice to remain stable. Also, if you have multiple PCIe slots you can try them out too, might be a faulty slot, also, try memtest and prime95 make sure it's not the memory or CPU that's causing weird stability issues.
 

emperor886

Reputable
Jan 5, 2015
3
0
4,510
Yep, done done and done (well, Aida64 instead of P95). No issues in any test/benchmark unless it uses the GPU. Even went as far as to run Intel Processor Diagnostic and ofc it was fine.

As for the 6 and 8 pin connectors, I have tried both from one rail, both from another rail, and one from each rail.

What I cannot for the life of me understand though is why F@H runs perfect when it has the card at 100% constantly. Does gaming require more juice than GPU computing?
 
In demanding games, your graphics card doesn't run at 100% all the time, it jumps around a lot, much faster than most programs can log, and for all intents and purposes it's running at 100%, it's the jumps that the PSU can't handle. That's why you'll see around here and other forums we always recommend getting the best PSU you can afford. I have a similar problem with my CPU. I can prime95 it for a week @ 4.8 and stay below 70C with no errors or stops, but damnit every time I run Diablo 3 it crashes. It's the only time it crashes, and it's just Diablo that crashes, so I leave it at 4.8 unless I plan on playing Diablo. It's not necessarily that it's a PSU issue, but an issue somewhere along the power delivery line. Could be the power through the PCIe lane, could be direct lead from the PSU. Don't take any of this as Gospel, I'm good with hardware, but there are people on here who are much much more knowledgeable than I am. This is just theory and speculation, not only with my issues, but with yours as well, you just hit a point where you blame voodoo and leave it at that. Time to get out the chicken bones.
 

emperor886

Reputable
Jan 5, 2015
3
0
4,510
Jossrik do you have C1E and C3/C6 states disabled? Might fix your Diablo problem.

So, jumps the PSU can't handle eh. Makes sense, F@H would be like a candle whereas gaming is like dynamite.

Problem is, under warranty Alienware will only replace parts with the same ones, not upgrade stuff, no matter how much proof I have. So I'll have to save up and deal with it for now. XFX Pro Series 1250W looks nice ^.^