2600K: thread loaded 100% all the time

[Falkentyne]

This is precisely what I said. This problem is *always* a hardware issue if it happens on a new install. I mentioned that this happened to someone else some time ago and it was faulty hardware. But I don't remember if it was board or CPU. Guess we'll find out. Board does sound more likely. Keep us posted.

[little_scrapper]

Its obviously that blackest of blackmarket bootleg copy of W7 that you are using. Try a legit disk. hehe

[[XC] Synthetickiller]

Its obviously that blackest of blackmarket bootleg copy of W7 that you are using. Try a legit disk. hehe

Love the joke , but this problem is super annoying.

Ewiz hopefully has a decent RMA turn around time. Claims 2 business days. I hope they are there today to recieve the RMA request since I just sent it.

Should be interesting. I'll report back later.

[[XC] Oj101]

That isn't software. Software spreads the load across all cores unless the affinity has been locked to one core. If you run SuperPi on a dual core CPU you'll see that both cores sit at 50%, which is equal to full load on a core. Windows juggles things around.

[[XC] Synthetickiller]

That isn't software. Software spreads the load across all cores unless the affinity has been locked to one core. If you run SuperPi on a dual core CPU you'll see that both cores sit at 50%, which is equal to full load on a core. Windows juggles things around.

I agree.

Locking to one thread is so odd. It affects 2 threads (1 core, 1 HT) in Prime95, but just sitting in Windows, only one is bogged down. Why 2K8 R2 isn't affected is beyond me.

I just hope the RMA process is fast (so far, it can take up to 2 business days to get one issued).

I'm tempted to order another one, but don't wnat to go to the hassle of selling the 2nd one on here.

It's probably just a fluke.

[danielkza]

That isn't software. Software spreads the load across all cores unless the affinity has been locked to one core. If you run SuperPi on a dual core CPU you'll see that both cores sit at 50%, which is equal to full load on a core. Windows juggles things around.

I'd say the same if the source of the load wasn't very specific: interrupts can be targeted at individual cores in a lower level than Window's scheduler, specially if they're coming from a defective driver/piece of hardware. I don't doubt the blame is on defective hardware/firmware, but it's very very unlikely the CPU is it.

[Falkentyne]

Im pretty sure the last time I saw this problem was also on this forums. Rather sure it was with a core i7, but I don't remember. May have been a C2Q. I just know it was hardware related, though. Maybe a search can find it..

[[XC] Synthetickiller]

I ordered a new board. The 2600k doesn't have this 100% thread spike. It must be the board. Flukes happen, case closed.

[Falkentyne]

Glad it's fixed. Told you it was a hardware problem

[[XC] Synthetickiller]

Now I'm fighting w/ my new board, LOL, go figure.

[[XC] Oj101]

I'd say the same if the source of the load wasn't very specific: interrupts can be targeted at individual cores in a lower level than Window's scheduler, specially if they're coming from a defective driver/piece of hardware. I don't doubt the blame is on defective hardware/firmware, but it's very very unlikely the CPU is it.

If it's a lower level than Windows scheduler can you really blame it on a virus or other piece of software running in the background?

--------

I'm glad you came right

[[XC] Synthetickiller]

I didn't even reinstall. My previous install booted up fine w/ the exactly same hardware (besides the boards). Just for everyone's edification.

[danielkza]

I guess we'll never find out what actually was causing the problem. Glad it's solved though

[[XC] Oj101]

A faulty board caused it?

[danielkza]

A faulty board caused it?

What exactly was faulty that caused a bug of such a weird nature? That's what I was wondering.