chipped athlon core

[pamf]

i accidentally bumped the heatsink and heard a crack, and this is what happened.. it still seems to work, but does anyone know if this might affect performance? random lockups, higher temps, anything? or did i just get really lucky?

if that link doesnt work try this ..

 

[Rahminator]

Run this combo for at least 24 hours: prime95's torture test, folding@home and distributed.net's RC5. From time to time, look at prime95's output and if it generated an error it will tell you and halt the process. In that case you will know there's something wrong and I strongly recommend you don't ignore it. Folding@home won't tell you if it found a mistake, but it stresses your system so much that it will reboot if there's something wrong in which case you have to take special course of action. RC5 is there just to increase the load. Remember, 24 hour burn-in period is the absolute bare minimum to see if your system is stable or not.

[Edit]Get motherboard monitor and look at the temps from time to time. Don't know if this is accurate or not, but I've heard that chipped cores increase cpu temp so monitor it carefully.

 

[pamf]

where can i get prime95? and 'folding@home' .. ive never even heard of that one .. i dont run rc5, but i do run seti, i suppose thats a good substitute? as for the temp, mbm says 58c .. that is a bit warmer than before, i used to run around 53c ;/ although some of that might be the thermal compound, since i had to clean it off to assess the damage and take a pic, might not have got it back on as well.. ill be damned if im taking that thing off again for anything less than a new fan. stupid things. ive got 486 heatsinks with superior clips than this stupid thing. at least i could take those off without a screwdriver.

 

[Rahminator]

Yeah, seti is a good replacement for folding so use that. Prime95 can be found here. When you install prime95 select the option that says that you want participate only in torture test. RC5 may be a lil confusing to a newbie so read the directions on how to set it up from the here.

 

[pamf]

newbie? pff. whats that? ..

FATAL ERROR: Rounding was 0.4999999554, expected less than 0.4
Hardware failure detected, consult stress.txt file.
Torture Test ran 3 minutes - 1 errors, 0 warnings.
Execution halted.

;/ now what..

 

[Rahminator]

If it fails after 3 mins, your system is very very flaky. Errors like that slowly but steadily corrupt your hard drive. Are you oc'ing? If you are then step down. My XP1600 at 1640 MHz used to throw a similar error after just a few minutes so I needed to step down. I ran it at 1612, but it rebooted after a few hours when running the combo. Then it was 1601 and it rebooted after about 13 hours. Now I'm testing 1589 MHz. Anyway, you're in trouble. It can be ram (to test ram, use memtest86 and run it for at least a few hours - 1 or 2 hours will suffice), motherboard/chipset or it's a sign that by chipping you simply damaged the chip.

 

[pamf]

i upped the core voltage to 1.80 (from 1.75) .. cpu temp is up to 61c, but prime95 has been running for 10 min so far with no problems .. i sure hope itholds up over night ..

never mind that. 12 minutes....

 

[Rahminator]

If you oc, go back to stock speeds and load fail safe defaults in BIOS. Run memtest86 to be absolutely sure it's not a ram problem. Check the voltage readings - 5V line really can't dip below 4.5V at full load and should be kept at >4.7 minimum, otherwise it's a PSU problem and you should get a replacement. If you took all these steps and it still errors out, looks like your cpu is shot. It could be the mobo, but the chance of that is very slim in your case.

 

[pamf]

if increasing the cpu voltage makes it go longer before giving an error, wouldnt this rule out memory, psu, etc?

 

[Spikesoldier]

61 degrees? Wow. I have my machines to turn off at 65.

 

[pamf]

it got up to 67 then began dropping, dunno why. back to 61. ive had this cpu running for weeks at 75c .. (back in the storied 'orb' days..)

 

[Wind]

<< if increasing the cpu voltage makes it go longer before giving an error, wouldnt this rule out memory, psu, etc? >>


I think tht CPU is on the verge of...failing. Use at as much as u can b4 it failed completely.

 

[pamf]

i sure hope not. i couldnt afford more than a cheap duron right now .. ;/

prime95 ran with dnet and seti in the background (at 1.85v) for a 26 hours, then i shut it down .. i hope it holds out at least another few months .. it's too bad my kt7a doesnt support xp's, the best i can get on this is a 1.4ghz tbird ;/

 

[microAmp]

Before you run your combo test and in case no one said anything I will (I briefly scanned the posts) bring all your settings back down to stock level, DON'T OVERCLOCK ANYTHING!!!!!! Then run your combo test. When you were o/c and thought it was stable it could have still failed the combo test!!!!!

My 2 cents......

 

[pamf]

if it passed overclocked, why run it at stock speeds? i plan to keep it overclocked, even if it eventually kills it ..

still i havent had any problems so im happy leaving it alone for now.. i wouldnt even care about the prime95 errors except that i run seti and i dont want to send back bad results..

 

[Jerboy]

<< if it passed overclocked, why run it at stock speeds? i plan to keep it overclocked, even if it eventually kills it ..

still i havent had any problems so im happy leaving it alone for now.. i wouldnt even care about the prime95 errors except that i run seti and i dont want to send back bad results.. >>



No errors at stock speed and errors at overclocked speed after a physical damage to chip is a good indiciation for you to back off and go back to stock speed.