Do I have the luckiest CPU in the world?

[Theo404]

So I decided to take the ihs of my san diego last night, and all seemed to go well, (btw for anyone who doesnt know what to do with the razor blade and chip sat it front of them, start on the corners and work your way in). I got round three sides fairly easily, but the last one was a little tougher to go through the middle of, this is why:



Thats a big old gouge going through several layers of tracks:



Look, you can see the chunk of PCB and copper left on the ihs:



AMAZINGLY, the chip seems to still work! Just passed 6 hours 45 minutes of small FFTs, and to boot I've dropped nearly 10C on my full load temp!



It loaded at 47/48C with the same ambient and voltage/frequecy, now full load after 6h 45m is just 38c!!

I was really suprised when the thing booted, what gives, how come it isnt dead??

[.sentinel]

WOW Thats what I would call luck. That is really close.

[5days]

I still think mine is luckier.

3 Chipped resistors and still going strong.
And its not little chips either, I nearly took a whole resistor off.

[Obi]

well back in the days, I had a duron 800MHz and when putting the heat sink on it I took a rather large chuck of the actual die off! then when I remounted the heat sink I didnt notice that the peice that had come off was still stuck in the thermal compund so when I tried to put the HS back on, it scraped the die even more! well with a peice chipped off the corner of the die and other parts scratched I booted and it worked just fine!

seems like the top layers are for show


insane

[DAK1640]

Incredible, you guys really take a lot of chances huh? Congrats on the CPU save and the significantly lower temps...

[EnJoY]

My 3200+ had 2 1/2 resistor packs taken off and still worked fine afterward.

[Theo404]

Resistors just fillter voltage, traces are a different story, I guess I just must of hit ground lines or something......

[kiwi]

I still think mine is luckier.

3 Chipped resistors and still going strong.
And its not little chips either, I nearly took a whole resistor off.

You mean prime and 24/7 stress test stable? That is a real luck

[Mr_Slinky]

yes congrats on the lucky save

[wwwww]

you guys are lucky...


why wont my 2.6C work

[Absolute_0]

Maybe if it survived having that tumor on it, and clocked to 3 Ghz on air, then it would be pretty lucky

[Onepagebook]

lol, that let me recall my old 0438 TPMW FX55 got lost 16 pins and even gain around 70 mhz IIRC

[Theo404]

Yeah, prehaps not the luckiest in the world, but it sure is REAL lucky for me.

I just cant get over the dropped degrees as well, 2.7xxghz was max stable with any voltage before maybe due to temp (it would load in the high 50s/low 60s if you bumped the voltage to 1.6xV), now priming at 2.8 with 1.47V, and I feel higher things coming.

[Thorry]

Jup remembers me of those old Duron Morgans too.

I had several which were chipped on all sides and still worked.

Once I had a chip which had already been abused by somebody, he gave it to me and I started to work on it. It was almost completely destroyed, not only had it missing corners but there was a big chunk missing in the center and a crack across the entire core.

It didn't work in any of my systems any more, so I gave it to my little brother. To everyones amazement the keyboard lights flickered a bit, then the whole system started beeping. After another reset the chip worked perfectly again, and even overclocked to 1500 mhz.

Later we had a competition who could get it's CPU the hottest whilst still running. The Duron topped out at about 95 degrees (measured on the heatsink), a Athlon 850 topped out at about 80 degrees and a Palomino 1600 ran an amazing 120 degrees before it crashed.

All chips worked perfectly after that, the duron ended up being snapped in 2 by a accident. The athlon 850 ended up getting 230 volts and exploded. The Palomino was used to test the maximum temp a chip could survive. It went into an oven at about 250 degrees for 1 hour and did not have any problems after cooling down. Then it went in about 300 degrees and still did not have any problems. But alas after being put into 350 degrees it died.

[5days]

You mean prime and 24/7 stress test stable? That is a real luck

Yup.

I dont really care for this CPU anymore though, I'm getting an Opteron soon anyway, and I dont think I'll get a reasonable price without 3 (1/2) resistors and an IHS. So it will be my practice CPU with DI (only in another 3 months or so though ).

I'll even try 2.0V hehe.

[Piotrsama]

The athlon 850 ended up getting 230 volts and exploded.

What do you mean?
You plugged the proc. directly to the wall electricity connector?

[Flib]

You should test if your SuperPI score decreased.
It could be that some Cache isn'T activated now, that would explain the lower temp!

[Daved+]

Real lucky guys doesn't scratch the PCB when they take the IHS off

[Theo404]

Real lucky guys doesn't scratch the PCB when they take the IHS off

Yeah, dont I know it!

[[cTx]Philosophy]

Im thinking of taking mine off this week but after reading this im thinking maybe not my X2 loads at 47 48 with dual folding going on..

Is there a good guide out there to prevent messups..

I remember reading back way back someone took ihs off and it looked clean as whistle. thats how i want mine to go :P

Looks like research time!

[brandinb]

You should test if your SuperPI score decreased.
It could be that some Cache isn'T activated now, that would explain the lower temp!

haha i dont think it works like that

the temp went down because better heatsink contact and removing two more thermal contacts and one layer of metal
________
Free drupal themes

[Theo404]

I think he meant the very large temp decrease, having less cache would make the chip run cooler, san diegos run hotter than venices....because of the larger cache. Also its very possible to kill cache by damaging a cpu like I have, so I thought it was a good point, however I took no screenshots of non-very overclocked super-pi, all I have are shots from trying out my home build phase change.