Ulz

Just a note: The way that's presented makes it look like there is a setting in windows blocking that, but it isn't. The game runs fine for a varying amount of time, sometimes seconds sometimes hours and then it just crashes. It's just a lottery honestly and it really annoys me because I just have to hope that THIS time I load it up it lets me play. I'm going to just buy a whole new set of memory and ditch this stick.

I haven't had a single bluescreen. Just every game crashes. And of course this:

I only just read this, but I ran memtest 86 and I got this: 2021-05-22 18:45:04 - MtSupportRunAllTests - Test execution time: 102.531 (Test 6 cumulative error count: 32450) 2021-05-22 18:45:04 - Number of errors exceed maximum error count of 10000 Not too good, and I initially thought it was memory related till it started mentioning that the "CPUs at fault" were 2, 4, 6, 8, 10 and that was out of 0, 2 ,4 ,6 ,8, 10 So I don't really know where to go from here. I'll attach the entire log if you want to look through it. MemTest86.log

I have tried some things on it. I ran windows memory diagnostics, I've tried different combinations of frequencies (which allowed me to discover that it's stable at 3200mhz, but I don't run it on that, I'm currently running XMP disabled because of my games being unstable) I'll run memtest, is there anything else I can do? I could also get some more ram right? I've also tried reseating it, I've tried it in every slot.

OK here's the problem. I have problems with my games crashing. All with the same errors: 887a0006 (DXGI_ERROR_DEVICE_HUNG), sometimes 887a0005 (DXGI_ERROR_DEVICE_REMOVED) (as in the card is REMOVED FROM THE SYSTEM) sometimes both at the same time in the case of cold war: 887a0005 (887a0006) or 887a0005 (887a0005) which is how I differentiate which errors they are, whatever is in the brackets. I have many crash logs, dxdiags generated by the game etc but it's driving me crazy. Warzone crashes, Cold War crashes EVEN FORTNITE crashes on low settings. Here are the specs: CPU: Ryzen 5 5600x (previously a 3200g) GPU: EVGA GTX 970 FTW 4GB (Previously an RX 580, and before that 2x GTX 770s) Mobo: ASRock B450 Steel Legend (on newest bios) PSU: Gigabyte P850GM 850W Modular 80+ Gold (Previously an XFX Pro Series 650w from 2013?) RAM: 1 DIMM of 8gb Corsair Vengeance DDR4 2400mhz Note that I have replaced MULTIPLE parts of this computer. I have tried EVERYTHING on the internet that relates to these errors, such as TdrLevel and TdrDelay + TdrDdiDelay and even as far as 0 filling all my hard drives and starting over. It all started with my RX 580 early this year. I was getting black screens causing the PC to reboot. I RMA'd it and got a refund as it was faulty under warranty. With this money I bought myself a GTX 970 and a second hand monitor (£180 + £40 in that order, good deal) during the obvious amount of downtime in which I didn't have a card, I had 2 GTX 770s on hand. This is where it gets complicated. One of them I already suspected didn't work as intended. They were both the same model, so I didn't know which was which as I put them both upstairs in a cupboard as a last resort spare if things went bad. During the time my 580 died, my friend's 590 also died, so I lent him one of my GTX 770s for use. It worked incredibly well and now that I think back on it, that left me with the suspected dying one. That one is also getting the quoted errors, and also in modern warfare (mainly shipment but sometimes warzone) it gets what I have dubbed the "christmas lights". You probably know what these are but I'll explain. At random intervals during the game, These little lights appear on the screen for a split second in a seemingly random colour. There are other issues such as the gunsmith not rendering correctly, the guns and attachments not loading and the guns becoming SO reflective that they turn into the sun and become a blinding white light that engulfs everything behind the UI. Anyway, my first GTX 970 arrived. This one had severe issues. For a while it worked OK. Still crashed a lot. Then it died. As in DIED. Screenshot: https://ibb.co/n1xr4Wg So I messaged the seller on eBay. He was confused, as the card was working perfectly before he sent me it. So he requested it back, and in return sent me my current card. I was very skeptical to use it as I didn't want any fault with this card to be on my end, so I bought myself a new power supply as this one was 8 years old. I installed it, did everything correctly on my end, and installed the card. It worked, momentarily, the games still crash and do NOT work well at all. Yeah, no artifacting or corruption but they still crash and basically no games are playable for an extended period of time, and I don't want to kill my card. What can I do to fix this, could it be CPU or RAM related, I dread that it's Mobo, I really cba getting a new mobo after less than a year. Please somebody help me, if I've left out any info or you have any questions please don't hesitate, I'll do anything to get a working PC again.

OK, that makes sense. It's an old PSU so it was time to upgrade anyway, that's just the final nail in the coffin for it I guess, given that you could assume it's killed 2 GPUs at this point.

OK, thanks for your help. I was just wondering, how come my GTX 770s can survive perfectly well but my 970 and 580 can't? I'm assuming it has something to do with the gtx 770 having a higher tdp but I really don't know.

It is a 650w XFX Pro Series, non modular, 80+ Bronze

OK, here's the context: I don't know how old the PSU is, I predict that it is almost 8 years old as the GPU's that came with this system were 770s. Since then I've undergone multiple upgrades. I upgraded my CPU to a ryzen 5 5600x, 8gb DDR4 Memory, ASRock B450 Steel Legend and an RX 580. That graphics card died less than a year after the upgrade (I bought it in June 2020 while the prices were good). I RMA'd the card and went back to my 770, no big deal. Recently (less than a month ago) I upgraded to an EVGA GTX 970 described as refurbished off an eBay seller. That card died in every way possible. It functions yeah but it's just barely hanging on (see picture). I just sent the card back to the seller and he swapped me for an EVGA GTX 970 FTW. It has arrived and I'm scared to put it in as I don't want to kill another card. This PSU has probably been in a lot of stress since I got it: I play ALOT of heavy games such as Cold War, Warzone, Far Cry 5 and other games that you could consider to be pretty power hungry. Before I played those, I was a streamer. I had 1 770 for playing and the other for nvenc. That could also contribute to a high power draw. Anyway, what do you think? Is my PSU dying? Is it safe to put this new card in? Have I provided enough information or do I need to give more for a verdict?

Your frame time is probably quite high, meaning it's probably a CPU bottleneck, although I can't see how it could be because it's still a really competent CPU for gaming. Check your bios settings, maybe you have XMP turned off for ram, as it is only 2666mhz. Ram overclocking isn't that hard and yeilded very good results for me, as I was having stuttering issues too. I was at 2400mhz, now I'm at 3200 stable all the time, which feels great on my mid-range hardware, however I'd say you're at a pretty high end, with the bottleneck being your ram. That's being nitpicky though, just try the ram, otherwise I don't know what to do.

That is AWESOME! Nice job, even though I'm almost a year late lol

well I just turned the voltage up to 1.35V and the clock speed up to 2866. Won't be as good as new duel channel 3200, but I'll have an improvement. Thanks for all your help I mean it, I was going crazy tryna figure it out, and I must have passed off the RAM issue so many times as just a minor issue even though I had told people that you need good ram to get the most out of ryzen. Have a great rest of your day from the UK!

150-350 Dollars. It's wide but eh, I'll come back when zen 3 comes out and figure out with everyone what the best is, I'm assuming it'll go with the trend and make Ryzen 5 the best Price per Dollar. Upgrading RAM first then going for CPU then case + fans.

1 8gb lol, did have an old DDR3 Rig with the same GPU's with duel channel, and I got the same performance So now I get it, the issue is 100% RAM lmao, thanks so much, that confirms my suspicions!

kmode_exception_not_handled refers to applications and drivers overwriting each others memory. Your error is most likely due to a corrupt driver. If you have a USB drive, remake your windows 10 setup USB and drop the drivers onto it in a folder afterwards, and DO NOT CONNECT TO THE INTERNET AFTER SETUP. Install all your drivers first, and then make sure it all works, then do the interweb stuffs. If it doesn't work, then it's hardware. Reseat your PCIe devices and RAM, and just check that all out. That's the basic troubleshooting but if that doesn't work, your stuff is defective. Also: don't forget these things: 1. That's the worst case scenario and 2. Google + Safe mode is your friend. It seems daunting doing troubleshooting but knocking things of the to-do list until it works feels great when it does work. Oh and I forgot, check your BIOS configuration