Faster after overclock is removed?

[1Wolf]

I've got an older machine (detailed below) that has been retired from being my "main" machine a year or two ago. It still gets used once in a while though. I built it originally in 2007 and then overclocked it to its current level (3.33Ghz) in 2008. Its been running at that speed ever since. It was never bullet-proof stable at that overclock. It was 24-hour stable in Prime95 with both "Blend" and "Large" FFT's but once every week or two....it would mysteriously reboot while playing a game or whatever.

A few months ago the stability got worse and it would mysteriously reboot about once a play session.

A few weeks ago the stability had dropped off to the point that it would reboot anytime a game was run.

At first I tried adding some voltage to get the overclock stable again but that chip never was a "great" overclocker to begin with so I didn't have far that I felt I could comfortably bump the voltage up. I bumped it up to just over 1.48 and, when it still was only stable 8 hours in Prime95 I was a little worried about bumping it further over the long term. The heat was still fine...it hit maybe 67*c under 100% load. But it still wasn't even close to stable in a game.

So I completely removed the overclock and set the thing back to its stock 2.4Ghz. Now games run fine. No crashing. No reboots. The thing is stable.

Weird thing is...its faster. I didn't benchmark it or anything...but it "feels" faster.

Why would 2.4Ghz be faster than 3.33Ghz?

Thanks!

System Stats:
Intel Q6600 G0 2.4Ghz overclocked to 3.33Ghz
EVGA nVidia 680i SLI motherboard
8 Gigs Mushkin
EVGA GeForce GTX 295

 

[MrGuvernment]

overclocking has always had a sweet spot, more and more and more Mhz doesn't always = faster performance in a sense as your entire system may not be well "tuned" together, memory, CPU and such

1.48 seem's dam high for a Q6600....or no been so long since i clocked one

Could just be your chip is finally dieing....

if it feels faster than something wasnt stable before...

 

[Spazturtle]

Unstable or near unstable overclocks slow the chip down.

 

[lilfiend]

I could be wrong but I thought those chips were good up to 1.55v 24/7.

If I were you I would have just backed off the OC a bit to ~3ghz if possible instead of jumping all the way to 3.33ghz from 2.4ghz

 

[Arcygenical]

My q6600 behaved similarly. After a few years at 4.0 it would get really slow in windows. Eventually icons would fail to load at boot - which always happened when i pushed the chip too far, or with too few volts - when i was attempting 4.1-4.5ghz on it at time of purchase.

Backed it down to 3.5 and it was butter smooth again.

A stressful oc causes all sorts of weird issues in windows.

 

[Spazturtle]

Just to point out, a new $35 1155 CPU performs the same as a Q6600

 

[cdoublejj]

I've got an older machine (detailed below) that has been retired from being my "main" machine a year or two ago. It still gets used once in a while though. I built it originally in 2007 and then overclocked it to its current level (3.33Ghz) in 2008. Its been running at that speed ever since. It was never bullet-proof stable at that overclock. It was 24-hour stable in Prime95 with both "Blend" and "Large" FFT's but once every week or two....it would mysteriously reboot while playing a game or whatever.

A few months ago the stability got worse and it would mysteriously reboot about once a play session.

A few weeks ago the stability had dropped off to the point that it would reboot anytime a game was run.

At first I tried adding some voltage to get the overclock stable again but that chip never was a "great" overclocker to begin with so I didn't have far that I felt I could comfortably bump the voltage up. I bumped it up to just over 1.48 and, when it still was only stable 8 hours in Prime95 I was a little worried about bumping it further over the long term. The heat was still fine...it hit maybe 67*c under 100% load. But it still wasn't even close to stable in a game.

So I completely removed the overclock and set the thing back to its stock 2.4Ghz. Now games run fine. No crashing. No reboots. The thing is stable.

Weird thing is...its faster. I didn't benchmark it or anything...but it "feels" faster.

Why would 2.4Ghz be faster than 3.33Ghz?

Thanks!

System Stats:
Intel Q6600 G0 2.4Ghz overclocked to 3.33Ghz
EVGA nVidia 680i SLI motherboard
8 Gigs Mushkin
EVGA GeForce GTX 295

i've been told by a tech whos been working on computers since the 70s that faster doesn't mean more performance. there is of sorts error correction in the cpu and or ram or something along those lines that help with stability, and at some point it's said that faster doesn't always mean FASTER.

i took it with a grain of salt but, i'd explain why it might be faster after UN overclocking.

also you could have been straining the cpu or cooling or causing something to throttle.

 

[BababooeyHTJ]

Oh, yeah. I had a similar issue. It wasn't that the core itself was unstable it was the mch or memory controller.

In fact if you're using 4 dimms I can assure thats what it is. I would be willing to bet that if you run Prime large FFT with all or your available memory set to be used it'll fail, quickly.

 

[chuckinbeast]

Oh, yeah. I had a similar issue. It wasn't that the core itself was unstable it was the mch or memory controller.

In fact if you're using 4 dimms I can assure thats what it is. I would be willing to bet that if you run Prime large FFT with all or your available memory set to be used it'll fail, quickly.

Interesting. I observed poor overclocked performance on a 2500k --- could that have been the culprit? In that case it was stable and more voltage ironed out the bad performance, but I wonder if your idea could have contributed. Was also using 4 DIMMs.

 

[BababooeyHTJ]

Possibly, but the op is using core2. Thats also where I ran into that issue. Its also much harder to work around on that P45. The extra dimms and the overclocked FSB really stress the northbridge.