I recently purchased a P4 2.53 GHz processor. I installed it in my new mother board and was surprised at the size of the heatsink but I assumed it must be ok because Intel sent it with the processor. The underside of the sink was not polished smoth but it had a piece of metal about the size of the CPU die attached to the bottom covered in what I assumed to be thermal grease of some sort. I also noticed that the metal patch was very soft, this I assumed would make up for the fact that the heat sink was not polished. I fired up the system and everything worked... I started installing SuSE and initially had problems with the video but overcame those by running in an all text mode. The installation was going smoothly but about 5 mins into it I started to get errors from the installer. I didn't think too much of this I just put it off to not having up to date drivers for some of my hardware and figured I could fix it later. Then, a Big red box appeared (Big red boxes are never a good sign) and the installation stopped. I hit the reset button still unsure what the problem was and my new MoBo started talking to me, "No CPU Installed" over and over again. Much head scratching ensued. I waited several minutes and then tried again, "No CPU Installed". I opened the case and removed the heatsink. I found that the soft metal had made a nice imprint of two corners of the CPU die and nothing else. Then the cold reality set in, I just spent over 400 CAD on a processor that cooked itself in under 10 min. I am now in the process of trying to return it to the retailer for a replacement. I do not wish to go through this a second time so I am asking for advice. I have been considering a water cooling solution, but I am haunted by the uneven imprint of the die on the heatsink. If the die is uneven will any cooling solution work? Can this unevenness be fixed with a liberal application of thermal grease (which is against the manufacturers reccomendations)?
"I found that the soft metal had made a nice imprint of two corners of the CPU die and nothing else."
that sounds like the heatsink wasnt sitting even on the cpu. I would be surprised if the cpu fried. I work for HP and seen systems run without the heatsink locked down. the fans just blew at full blast. the next time u install the heatsink take off all the grease and put on some artic silver compound and make sure it is sitting level. After putting on the heatsink but before u lock it down try wiggling the heatsink to make sure its level. then lock it down.
Too much grease is just as bad as no grease. You only need a little to fill in the microscopic imperfections in the die and heatsink. You want metal to metal contact.
Hope this helps.
P4's are not <i>supposed</i> to fry. That's the single major point which intel fanboys like to point out to AMD users. THG did a video showing actual removal of heatsinks, and the P4 Doesn't even <i>crash</i>.
This may seem a silly question, but are you sure there wasn't a little plastic strip on the thermal stuff on the bottom of the Heatsink? If there was it should have been removed.
I can only suggest that you try cleaning the cpu/heatsink thoroughly, and use some Arctic Silver and re-seat the Heatsink.
I grounded my self to the chasis of the box. I opened the packaging for the CPU inside the box so my body was in constant contact with it creating a zero potential environment. I picked up the CPU and placed it on the socket and plunk it fell in I closed the lever and completed the electrical connection to the chips ground pin. I beleive this was an over cautios approach. Also the CPU did work for a while, I would have expected zero function if it had been fried by a static discharge. The fact that it ran for a while is what made me think it was a thermal problem. Truthfully I have been a hobbyist robot builder for nearly a decade, I have used CMOS and TTL chips extensively and have never cooked one with a static discharge. I have even forced some advanced cmos tristate buffers to operate at 3 to 4 times their rated current with out having them fail (I have even given my self burns that blistered doing the old touch check to see if the chip was running hot). I will absolutely agree that I am comparing stones to potatoes when I compare a 20 pin dip to a socket 478 cpu but the static precautions are the same regardless, wouldn't you agree?
Apart from Metal's suggestions above, not at all, but if it was simply a case of thermal death, it would be the first confirmed case I've ever heard of, and it would <i>really</i> annoy some Trolls around here, because they would have their soapbox knocked out from under them.
Of course this doesn't help you though. Sorry about that.
Several questions. First, I just ordered P4 2.5 Retail (I couldn't find an OEM anywhere), will the thermal compound and heat sink already be attached to the processor? If so, how do I remove them because I also ordered Arctic Silver III and a Volcano 7+ hs and fan? Finally, what is the best method of applying the Arctic Silver?
Sounds like you protected yourself pretty well to me. As I said, it may seem like a stupid question. (Though I have even known a pro once who forgot for one quick moment. That was all it took though. Heh heh.) However, I was actually more worried about a small discharge to the motherboard than I was to the CPU. I've seen that and/or tiny scratches made with screwdrivers that often have unusual results where it seemed fine for a while, but not for long.
Heck, I've even seen a slightly bent case keep an AGP card from seating correctly so that the system crashed every time someone walked by.
Anywho, in your case it sounds like something was seriously wrong. I've never heard a single case of a P4 dying a thermal death. So it's entirely possible that the CPU had been defective before you even got it. Still ... I'm not so convinced that the motherboard isn't the culprit. And I'm afraid that I'm really confused about the heat sink problems...
When I used to work in a shop I only grounded myself because that was the only thing my boss could ever yell about (he didn't know jack).
At home I've drug my socks over the carpet carrying a mobo and video card without the bags. I've slapped cards in the case while it was still plugged in. Plugged the monitor into the video card while it was still on. Pulled on the hose of my watercooler too hard and watched it spray (the power was off here luckily but the pump was going). I've done many more horrible mistakes with power supplies as well. The only thing I never mess with is the inside of a monitor. That thing will knock you on your ass or kill you if your not careful. (I know this from experience)
But I have never, ever, fried anything from being careless.(Except 2 Tbirds and a crushed Duron but those dont count, they always fry, ohh and my finger on a P2 which had it's heatsink mysteriously fall off)
Where theres a will theres a way. People can fry anything nowerdays :smile:
Heh heh. No kidding. Luckily I haven't ever fried anything myself (well, not counting the occasional egg or whatnot) yet, but then, I always try to keep myself grounded when working on a PC.
Of course, I have seen power surges in the phone line fry a few modems. That's always no fun. Somewhere in my computer graveyard is a stack of dead modems. Heh heh. I guess that's what you get for living out in the middle of nowhere for a while, is bad phone wiring.
Well, after reading all of your responses I think I can conclude that the CPU is probably not the culprit. Some research has led me to a new possibility, My PSU is not providing enough current to the mother board. I only paid 60 CAD for my case and the PSU was included. I thought this was a good deal, why would some one make a PSU that could not do the job for which it is intended.... To make more money!!! I have sent an enquiry to ASUS vis a vis the P4s533 mo bo (sis 645). I can not find any specefics on the web with respect to the design envelope for voltage and current for optimal operation.
My next task is to find out the specific envelope for voltage and current and measure the two PSUs I own and if they do not pass muster I will buy a new PSU that does. If that doesn't solve my problem I will have to find a kind PC vendor in St. John's who will test my CPU and MoBo with a different MoBo and CPU respectively. If all efforts fail I will exchange my CPU and MoBo for new ones from the vendor.
As long as I have your attention, I have PC3200 ram from corsair cas 2.0, any body feel strongly about a specefic MoBo?