Computer Crashing while gaming. I think my GPU may be fried. I need help :(

Status
Not open for further replies.
OCZ is good. I used them before. I like their RAM, too. If it gives you trouble, you can always RMA it. I had to RMA a few memory chips. Quick easy service.
 
Well, it's a good idea to upgrade the PSU, guess I screwed up in this thread to, I wasn't paying attention to everest as I haven't cared nor accepted them as a company since they removed the home version...

But, the VRM seems to be causing some serious issues, and could end the end, it self fry if you don't take action now. The reference design of your card, had them heatsinked, and covered by the GPU/DRAM heatsink it self. BUT, manufactures like to cheap out on heatsinks on all parts, just like your GPU, motherboards have a set of VRM's for the processor, and a lot of ~$50 motherboards, even some higher end ones, don't have a heatsink on them, and these babys, get HOT...

But, here is a picture to help you find the VRM, just be carefull to not short out the contacts on them, use a non-conductive THERMAL PASTE GLUE, or THERMAL TAPE, I would go with thermal tape if I was new at adding small heatsinks like that, just because it's a safety type of deal...

VITEC = VRM chip, http://resources.vr-zone.com/yantronic/1950pro/thermalpower/2phase.jpg
Atleast, I think it's the VRM, I know it's in circuit wit a transistor bank... Also, in that picture, are transistors, with a black piece covering them up, those I know for a fact get incredibly hot, and in the reference design, are covered by the main heatsink, but, it seems a lot of manufactures decided to safe ten cents and leave it off of those, so.


Also, here is the datasheets for your VITEC chip, note, it is designed to handle 125C, http://www.viteccorp.com/data/af4268.pdf

Also, here is a good read on your GPU, seems the "non branded" version does have an issue with the chip I pointed out.
black screen after playing games for a while (Radeon X1950Pro) [Archive] - techPowerUp! Forums
Two things came up.
Power supply and the vItec chip peach pointed out.
Placing one of the spikey Zalman heatsinks on the vItec chip apparently worked.



So, if you can, get some copper zalman heatsinks with thermal tape, or if you have experience, and know for a fact that you wont leak/squeeze out thermal glue, get some high quality stuff from newegg and use that, but be warned, a lot of the times, you can short contacts out using thermal glue if it's conductive when cured, so look for non-conductive when cured and let it cure. You should then be set to go.
Get pure copper, no painted/silver or anything like that, you should use a pure copper heatsink with fins on it, and if possible, have a fan installed blowing air towards that card to help remove the heat.
BTW, once again, sorry for ignoring a previous picture, and for this long *** post.
 
I purchased and installed the PSU that I linked. The GPU VRM is getting even hotter now when I run a gpu stability test. The temperature reached 124 degrees C.
I turned WoW on for a few minutes to see if the problem recurs, for about 5 minutes it seemed fine. Then I turned off Vertical Sync, and my FPS jumped to around 200. I got a black screen for a few seconds, and then it came back. I immediately re-enabled Vertical sync. I won't be trying that ever again.

Great post, corrosive. When I was picking up my PSU, I tried to find the heat sinks you mention, but there were only moderately sized blue ones. And I only see one chip that resembles what the picture shows. It's near the fan on the GPU (near where the PCI-E power cable plugs in). It's extremely tiny. Around the same size as a pinky-finger nail. It does appear to have a black pad-like thing adhered to it. Maybe there are more of these chips underneath the current cooler that is on the GPU? I don't know... I'm going to go back to the store tomorrow maybe and see if I can find some heat sink that will fit, and I may change the location of one of my case fans to blow directly onto the VRM. I may just buy a new card.

And lol, you don't have to apologize. If anything, I should be apologizing to you guys :). Thanks you guys for all of your help.
 
Yea, the VRM has ALL the electricity your GPU will use going through it, so it can get REALLY hot, and 125C is the max the VRM can handle safely when operating...

The reason it is running hotter though is because you have more power available for the card to use, so, the PSU before was somewhat, degrading performance, but, possibly in the end saving your cards life... The blue ones will work if your desperate, or, if you can hold off and keep an eye on it, these will work great.
Newegg.com - Computer Hardware,Fans & Heatsinks,Memory & Chipset Cooling,Copper

Just grab the right ram sized heatsink, that don't have clips/fans or anything like that, and that would fit your VRM chip, Just to note, larger doesn't matter, as long as the card can go back into place, and the heatsink doesn't contact other chips/parts.
 
Ok, cool. I'm just gonna go pick up the blue ones. Should I take the current cooler off of the card and clean it out thoroughly? I've never used thermal compound or anything before, but I'm pretty confident that I could do it without messing anything up too bad.

One other thing..I'm not exactly sure where the VRM are on my card. I only see one of those chips, and I'm not even sure if that's what it is. I don't know if there are more or something. And if they all have that black pad-like thing glued onto them, should I remove them and re-seat them or something?

You have been a great help man, thank you.
 
The majority appear to be covered up by the main heatsink on MOST cards. Also, a good blast of compressed air going in backwards through the heatsink will help blow the dust out, just use tweezers to pull the chunks out when it gets to the fan... -.-

Unless you feel like learning how to use thermal paste. :3

I know the transistors get incredibly hot, and they are generally covered up by heatsinks, that is what's under that one black piece, the VRM is the chip that has VITEC from what I have gathered.

Here is a picture I edited while in the middle of researching your card, VR = Voltage Regulator, the M stands for Module, which is why I think the VRM is the VITEC piece, as it is a VRM, but, a VR is also the transistors covered under those pieces of copper... Note the placement of that black piece, it's right where a heatsink in the REFERENCE design is, the reference design is the design ATI says will work properly for the card, manufactures seem to just wanna save a few bucks and hack off an ounce or two of copper.

card8.jpg
 
Microcenter has a good sale on some new cards, so f*** it. I'm buying one lol. It's about time to upgrade. I can get a new card for almost the same price as some new cooling equipment. I will post up what I purchased when I return from the store.
 
Status
Not open for further replies.
Back
Top Bottom