I sold my friend a 3.6 Q6600 (rig)...

[daveybrat]

Seriously, just clock it down to 3.0GHz and call it a day. He'll never even know the difference in speed for what he's doing.

3.6GHz is quite high for that old puppy.

 

[tweakboy]

This.

2) There is no reason to have it overclocked for the activities he is using it for.

Oh really,,, Just remember the most notice you will notice in speed is a SSD. end of story.

What are you planning on getting.

Treat yourself to a i7 CPU .. gl

 

[VirtualLarry]

Are the crashes heat related or not?

He said that this latest one was not, and that he did not have the PC enclosed. (That was the first thing that I asked.)

Way back when he first got the computer, he did say that he thought that it probably overheated, because he had the PC enclosed. It has not been enclosed since.

I just hope nothing got damaged. I don't have the money to refund him, and I don't give refunds on PCs anyways. But I'm more than happy to give him an even/greater swap for another PC I have in my "stable". Like a Thuban 1045T machine.

 

[Emulex]

WTF!? taking the case off is the worst possible thing you can do !! Fluid dynamics majors get in here and explain!

 

[VirtualLarry]

WTF!? taking the case off is the worst possible thing you can do !! Fluid dynamics majors get in here and explain!

It's in the case, when I said it wasn't enclosed, I meant that the case wasn't in a cabinet or something.

 

[Idontcare]

Had many Q6600s myself, I kept them clocked at no higher than 3.3GHz though.

3.6GHz is not THE problem for G0 kentsfield quads, THE problem is the chipset/mobo and the need for pushing the FSB to 400MHz in order for the Q6600 to hit 3.6GHz. It is very hard on the mobo to push the FSB at that speed with a quadcore and its associated demands on the power quality. So much easier (and sustainable) for a dual-core than a quad.

So when you describe the issue you are describing, my first suspicion is the mobo, that the 400MHz FSB has pushed things too far for too long and it is no longer able to reliably operate at that FSB speed.

 

[VirtualLarry]

Had many Q6600s myself, I kept them clocked at no higher than 3.3GHz though.

3.6GHz is not THE problem for G0 kentsfield quads, THE problem is the chipset/mobo and the need for pushing the FSB to 400MHz in order for the Q6600 to hit 3.6GHz. It is very hard on the mobo to push the FSB at that speed with a quadcore and its associated demands on the power quality. So much easier (and sustainable) for a dual-core than a quad.

So when you describe the issue you are describing, my first suspicion is the mobo, that the 400MHz FSB has pushed things too far for too long and it is no longer able to reliably operate at that FSB speed.

I really doubt that's the problem, it's on an x48 board that is specced out for 1600 FSB.

 

[Ketchup]

VirtualLarry, he probably neglected to tell you he smokes 2 packs a day in that room.

 

[VirtualLarry]

He said "I am happy with everything the way it is". I guess crashes and all. Well, you can't say I didn't try to rectify the situation. His warranty on the PC runs out at the end of the month.

I guess if it crashes every two months or so, then that's not so bad. If it were crashing every day on him, he might be singing a different tune.

Personally, my standard is basically zero crashes.

 

[Bill Brasky]

Seriously, just clock it down to 3.0GHz and call it a day. He'll never even know the difference in speed for what he's doing.

3.6GHz is quite high for that old puppy.

This is what I'd do. He'll never notice the difference and it won't crash anymore...

 

[VirtualLarry]

http://dalehyde.hubpages.com/hub/Google-Chrome-Crashes-In-Windows-7

I think I'm going to point him to this page.

 

[tweakboy]

Probably both!

 

[tweakboy]

Had many Q6600s myself, I kept them clocked at no higher than 3.3GHz though.

3.6GHz is not THE problem for G0 kentsfield quads, THE problem is the chipset/mobo and the need for pushing the FSB to 400MHz in order for the Q6600 to hit 3.6GHz. It is very hard on the mobo to push the FSB at that speed with a quadcore and its associated demands on the power quality. So much easier (and sustainable) for a dual-core than a quad.

So when you describe the issue you are describing, my first suspicion is the mobo, that the 400MHz FSB has pushed things too far for too long and it is no longer able to reliably operate at that FSB speed.

I was running 463x8 thats 3702Mhz stable, not stable anything above this. gl

 

[Matt1970]

I agree with the downclocking suggestions.

 

[VirtualLarry]

Well, I had the opportunity to talk to this friend, and he brought up the PC, said it "froze" on him like 20 times in the last month or two. Also mentioned boot-loops, after the freezes. Said that if he unplugged it and left it for 10 minutes, it would always boot up no problem, though. (To me that sounds like a thermal issue? Maybe the PSU?)

I asked him if he was monitoring temps, he showed me CoreTemp readings of up to 69C, while his PC was watching a twitch stream, having an intensive web page open, and on Skype.

I thought that was possibly a tad warm, but not warm enough to cause crashes.

He said that he had removed the center fan out of the Tuniq Tower 120 and cleaned it.

At this point, I think I sold him the PC at least 2-3 years ago.

I told him to go into the BIOS and drop the front-side bus (host clock) to 333 from 400. I don't know if he is capable of doing that on his own though.

I'm quite willing to help him get it stable again, if he brings it over, but he's interested in buying a new one.

Would an IB 3.4Ghz i3 be faster than a Q6600 @ 3.2? (Let's assume that the current 3.6 speed is no longer stable.)

 

[teejee]

It is very common that OC headroom gets a little worse after a while. Nothing to worry about, just lower the clock. 3.3 Ghz should probably be OK.

 

[Flapdrol1337]

X38 and X48 northbridges get unstable if they get too hot. A quad at 400 fsb is pushing it. Maybe the thermal paste on the northbridge needs replacing, or, just run at lower fsb, point a fan at it, etc.

I have an X38 system that refused to boot a month ago, tightend the screws on the northbridge heatsink and it works flawlessly agian.

 

[cbn]

Would an IB 3.4Ghz i3 be faster than a Q6600 @ 3.2? (Let's assume that the current 3.6 speed is no longer stable.)

Here is one of my favorite articles for comparing Ivy Bridge to Core 2:

http://www.tomshardware.com/reviews/ivy-bridge-wolfdale-yorkfield-comparison,3487.html

 

[Spungo]

This.
I am never selling on old hardware to familiy or friends(anymore) .. youll be on call 24/7 for the rest of the hardwares life.

Which is really weird because nothing else is like that. You would probably hit me if I bought your jacket then called you repeatedly when I ripped it or left it on the bus.


In general, NEVER overclock for someone. You can suggest people overclock, but never do the overclocking. Give a man a fish and you feed him for a day. Teach him to fish and you feed him for life. Give him a fish that isn't cook properly and he'll die.

 

[Denithor]

Larry, if you're going to necro a thread, mod the title to reflect it...

20 freezes in a month, followed by boot-loops that clear up after idle time, sounds to me like a thermal problem. But honestly, at this point, I think you're better off just moving him to a newer system. If you fix this one, it's going to continue limping along and giving nothing but headaches for you to deal with as it continues to get older and older.

 

[Engr62]

...Said that if he unplugged it and left it for 10 minutes, it would always boot up no problem, though. (To me that sounds like a thermal issue? Maybe the PSU?)...

This sounds exactly like a power supply problem I had a while back. The PSU would reset after it was unplugged for a few minutes, then no problem until the next random shutdown. I thought mine was heat related as well, but it just turned out to be a crappy PSU. Replaced it and never had another problem with random shutdowns.

 

[rbk123]

I'd wager it needs a clean OS install. I'd bet he's not skilled at installing software and there's all kinds of crap on it now.

My main rig is still a Q6600 running at 3.4G for 5+ years (at least, maybe 6+), but I've reloaded the OS a few times as I went to Win7 from XP and then to a couple of different SSD's.

 

[Spungo]

This sounds exactly like a power supply problem I had a while back. The PSU would reset after it was unplugged for a few minutes, then no problem until the next random shutdown. I thought mine was heat related as well, but it just turned out to be a crappy PSU. Replaced it and never had another problem with random shutdowns.

As I get older, I'm starting to think computers are more magic than anything else. Any crash can be caused by anything. For example, running Folding while doing something 3D will crash my computer, but Folding alone or gaming alone will not. Solution? Increase the CPU voltage. Seriously. This actually worked. My computer doesn't crash anymore.

Regardless of how to fix the problem, you can start searching for the problem:
-Run OCCT small units to test the CPU alone. IMO, this causes the fastest CPU failure. I love OCCT because it makes a weird quacking noise on failure

-If that passed, run OCCT Linpack 90% memory to test overall CPU/mobo/ram. I've never actually seen Linpack fail at something that is OCCT or Prime stable, but others claim it can. This is probably because my northbridge, memory, and HT are always underclocked slightly.

-If that passed, run the OCCT GPU test to test the video card's processors.

-OCCT's video memory test doesn't work on AMD cards, so for those I use Furmark. Test the video card memory by setting Furmark to at high resolution, full screen, with the highest antialiasing. If your video card is screwed, this will crash almost immediately by the screen artifacting then crashing to blue screen. If it crashes to a black screen, it's a PSU failure.

-If all of those passed, run the OCCT PSU test. This will crash to a black screen if the PSU sucks.

Even when the tests complete successfully, look at the voltage graphs OCCT saves. A good PSU will show sharp, right angle voltage changes. A really screwed up PSU will show wavy voltage changes, almost like the PSU is saying "yeah I'll change the voltage, just hold on a minute."

Like I said in the first paragraph, try repeating the tests while Folding GPU is running. It seems like Folding and gaming use two completely different parts of the video card.

 

[gmaster456]

Remove the OC. For what he does I doubt he'd notice anyway. It's an 8 year old CPU you can't expect to push it like that.

 

[2is]

This. Q6600s were known to be easily overclockable, but for a chip that old, 3.6 is pushing it a bit too hard Imo.

Even brand new, 3.6 was not childs play. I build many Q6600 based machines and only one hit 3.6 with a relatively little amount of effort. Some hit it with a lot of tweaking and a lot of voltage while others didn't get past 3.4GHz no matter what.

I don't think the OP is on the hook for any sort of guarantee though, if the transaction happened 2-3 years ago. But in the end, yeah, 3.6 is asking for a lot considering the age. Also, my experience with these systems is that the motherboards power delivery starts degrading before the CPU does.