Associate
After testing it at stock ang setting all the right stuff in bios, i pretty much jumped to 3ghz almost, though its a problem it if doesn't work, as you can't be sure where the problem is.
-
Competitor rules
Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.
***The Official Q6600 Overclocking Thread ***
- Thread starter easyrider
- Start date
After testing it at stock ang setting all the right stuff in bios, i pretty much jumped to 3ghz almost, though its a problem it if doesn't work, as you can't be sure where the problem is.
Associate
No not at all, 3GHz is stock speed for pc-5300, 1333FSB is stock for all new P35 boards and pretty much all G0's will run 3GHz at there stock voltage.
Associate
It's just a bit poor that there are such huge variations in the quality of the chips. I read a forum where several other people had got their Q6600's from the same place in the same week, each having crap VID's and core temp problems. Though as it runs stable at stock, nothing can be done about it 
Associate
That’s just how it goes, you are never guaranteed a good overclocking processor, you can just increase your chances greatly by plucking a cpu from a batch of known good overclockers.
Its been like that since I remember, back in the days when I used to overclocked amd thoroughbreds and probably way before I got into this stuff
Its been like that since I remember, back in the days when I used to overclocked amd thoroughbreds and probably way before I got into this stuff
Associate
When i get home tonight i'm going to lap the cpu properly (can't believe i even tried it on the Zalman HS, it has a mirror finish lol, but only did it a bit, hopefully it hasn't damaged it), because i'm not happy with the core temp variation. Also didn't actually clean the grease from the first application off properly, easily adds a couple of C. Got some Multi-purpose PC wipes from work, not strong alcohol base, but best i could find, i'll just have to dry it well.
Then i'll spend a good 15 mins reseating the heastink perfectly without budging it into place. Whatever i get then i'll accept. Though i'll wait a few days before OCing high again, as that grease takes several days to bed in.
Then i'll spend a good 15 mins reseating the heastink perfectly without budging it into place. Whatever i get then i'll accept. Though i'll wait a few days before OCing high again, as that grease takes several days to bed in.
It's just a bit poor that there are such huge variations in the quality of the chips. I read a forum where several other people had got their Q6600's from the same place in the same week, each having crap VID's and core temp problems. Though as it runs stable at stock, nothing can be done about it
We can't mention competitors, and I'm only asking where you got yours for point of refernece - as I think mine came from same place.
Did it begin with an "S" and is it a peripheral for a PC that was commonly used in the 90s but they normally aren't sold idividually anymore?
thanks?
Just curious as I may get another one but if the VID is the same in batches and not totally random then I will prolly get one from OCUK instead this time and sell the one I have.
At 3.5Ghz now, but needs 1.464v....
I think you may be being a little optimistic, from reading most of your posts it just seems that you have a very mediocre chip like mine. The 10c temp difference could well be resolved from lapping, but it doesn't neccasarily mean you're going to get a good oc. I really can't see 3.6Ghz being attainable, not meaning to be rude at all
But will be interesting to see your results, who knows.
Just lap the CPU down to the copper using 400 grit, then work your way up to 1200.. all done on a flat (eg, mirror) surface. Once thats done, if your cooler does have a mirror finish, then reseat.
I'd personally whack around 1.5v through it to see how high itll go.. probs go straight for 400x9 with the memory at completely relaxed timings and see what the temps looks like and if it can run prime for a bit. I'm sure it'd do 3.6ghz on water no problem.
I'd personally whack around 1.5v through it to see how high itll go.. probs go straight for 400x9 with the memory at completely relaxed timings and see what the temps looks like and if it can run prime for a bit. I'm sure it'd do 3.6ghz on water no problem.
Associate
Hernaldo: Yes you go it, all a bad batch, though saved like £20 heh. They were the cheapest OEM.
I've also read several other people getting them from the same place in the same time period, i wouldn't recommend getting another from them any time soon, which is a shame because the service is commendable. And no, the 10C difference remains
Uzi: I just finished lapping the CPU, using only 1200 grit. It took until now from when i got home from work at 6:30. It is a brilliant mirror finish, however, it proved how completely concave it was, as it took a long time before i even reached the central area. Surprisingly, the Zalman was also slightly concave, so i cave that a quick lapping, but not as thorough. I didn't go down to the copper on the CPU however, jsut until it was mirrored pretty evenly.
What i did notice is 3 specs on both the heatsink and CPU, almost like grit embedded in. No amount of lapping removed it, however, the surfaces are prefectly uniform.
The Zalman thermal grease i painted pretty much perfectly onto the CPU only at worst i may have a tiny tiny bit too much, but Zalman also recommends application to both surfaces.
Temp wise, well my hottest core is 32C at the moment OC'd at 3ghz. CPU reads 18C in Asus probe as reference. I'd say its dropped a couple of degree's. I dropped my vcore to 1.35 (1.33 due to vdroop). That's pretty much my VID, so i'll do some OCCT tests and a few quick small FFT's and let you guys know, under load is the proof of the pudding...
I've also read several other people getting them from the same place in the same time period, i wouldn't recommend getting another from them any time soon, which is a shame because the service is commendable. And no, the 10C difference remains
Uzi: I just finished lapping the CPU, using only 1200 grit. It took until now from when i got home from work at 6:30
. It is a brilliant mirror finish, however, it proved how completely concave it was, as it took a long time before i even reached the central area. Surprisingly, the Zalman was also slightly concave, so i cave that a quick lapping, but not as thorough. I didn't go down to the copper on the CPU however, jsut until it was mirrored pretty evenly.What i did notice is 3 specs on both the heatsink and CPU, almost like grit embedded in. No amount of lapping removed it, however, the surfaces are prefectly uniform.
The Zalman thermal grease i painted pretty much perfectly onto the CPU only at worst i may have a tiny tiny bit too much, but Zalman also recommends application to both surfaces.
Temp wise, well my hottest core is 32C at the moment OC'd at 3ghz. CPU reads 18C in Asus probe as reference. I'd say its dropped a couple of degree's. I dropped my vcore to 1.35 (1.33 due to vdroop). That's pretty much my VID, so i'll do some OCCT tests and a few quick small FFT's and let you guys know, under load is the proof of the pudding...
Associate
Results: (To be updated) (VID = 1.325v)
@333mhz x 9 FSB 1:1 (3Ghz)
Idle (browsing): Core 0 = 33 / Core 1 = 28 / Core 2 = 22 / Core 3 = 27
100% load Peak: Core 0 = 55 / Core 1 = 52 / Core 2 = 47 / Core 3 = 51
Vcore = 1.33 (set to 1.35 BIOS)
@360 x 9 FSB 1:1 (3240Mhz)
Idle (browsing): Core 0 = 33 / Core 1 = 28 / Core 2 = 22 / Core 3 = 27
100% load Peak: Vcore = 1.33 (set to 1.35 BIOS) = Crash in small FFT / Vcore = 1.344 (1.3625) = Crash in small FFT.
So tried:
@399 x 8 FSB 1:1 (3192Mhz)
Idle (browsing): Core 0 = 33 / Core 1 = 28 / Core 2 = 23 / Core 3 = 27
100% load Peak: Core 0 = 57 / Core 1 = 55 / Core 2 = 49 / Core 3 = 54
Vcore = 1.344 (1.3625)
Interesting, the thing holding me back on this system is the least i'd expect...the overclocking monster Q6600.
Seems by turning the multiplier down, i become stable under stress test, using the exact same voltages that crashed me on x9 multi. My Motherboard, i think, can handle upto 520mhz FSB. This will take some serious trial and error.
@333mhz x 9 FSB 1:1 (3Ghz)
Idle (browsing): Core 0 = 33 / Core 1 = 28 / Core 2 = 22 / Core 3 = 27
100% load Peak: Core 0 = 55 / Core 1 = 52 / Core 2 = 47 / Core 3 = 51
Vcore = 1.33 (set to 1.35 BIOS)
@360 x 9 FSB 1:1 (3240Mhz)
Idle (browsing): Core 0 = 33 / Core 1 = 28 / Core 2 = 22 / Core 3 = 27
100% load Peak: Vcore = 1.33 (set to 1.35 BIOS) = Crash in small FFT / Vcore = 1.344 (1.3625) = Crash in small FFT.
So tried:
@399 x 8 FSB 1:1 (3192Mhz)
Idle (browsing): Core 0 = 33 / Core 1 = 28 / Core 2 = 23 / Core 3 = 27
100% load Peak: Core 0 = 57 / Core 1 = 55 / Core 2 = 49 / Core 3 = 54
Vcore = 1.344 (1.3625)
Interesting, the thing holding me back on this system is the least i'd expect...the overclocking monster Q6600.
Seems by turning the multiplier down, i become stable under stress test, using the exact same voltages that crashed me on x9 multi. My Motherboard, i think, can handle upto 520mhz FSB. This will take some serious trial and error.
Last edited:
Associate
If you both got your Q6600's from the place I think you did (is it the letter S and another name for a tin?), then that's where mine came from, although I did get the retail version. Maybe it is the OEM's that were a bad batch, because that e-tailer hasn't got the OEM version listed anymore.
Anyway, i've now got a nice new ultra 120 extreme from OcUK sat on top of my Q6600
It's taken the temps at 3.2 (1.25vcore) from about 68-70 with the stock cooler to 46-48 with p95 small ffts
Temps should drop a bit more over the next few days as the AS5 beds in.
Now it's time to push this cpu a bit - not too much yet though. Got to let the AS5 settle first.
Keep us posted on how you're getting on v3g3tto. Hopefully you'll get to 3.6!!!
Anyway, i've now got a nice new ultra 120 extreme from OcUK sat on top of my Q6600
It's taken the temps at 3.2 (1.25vcore) from about 68-70 with the stock cooler to 46-48 with p95 small ffts Temps should drop a bit more over the next few days as the AS5 beds in.Now it's time to push this cpu a bit - not too much yet though. Got to let the AS5 settle first.
Keep us posted on how you're getting on v3g3tto. Hopefully you'll get to 3.6!!!
Associate
Dunno whether this is a silly question or not, but i was under the impression DRR2-800 would overclock? Tried 850mhz, basically so i could keep my voltage low by using a x8 multi, 425 FSB, but it doesn't post and saus overclock failed.
Hmm...i know for a fact this RAM goes to 900mhz, i tried loosening the timings to 5-5-5-18 too but same.
Anyway, after trial and error with voltages, i'm upto 3.4ghz with 1.4125v (or 1.39 due to droop). Temps at idle are 2-3C higher across the board, acceptable. Dunno if i wanna try the small FFT's, i know it'll fail at such a harsh test, but gotta be done i guess...
...to be honest, this chip cost me £152, it's not a huge loss as such if it's a dodo
Edit: RAM has hit a wall at 820mhz, so at x8 multi, thats around 3.2ghz, CPU has also hit a wall at 3.2ghz, at 3374mhz now, i get an error on small FFT's, and OCCT crashes.
what..the...hell. This is with 1.425v now. How can both my RAM and my CPU be dodgy, they jsut won't OC well and i've ran out of idea's. How can that ram, which i know can be pushed WELL beyond 900mhz even at 2.1v, not barely pass the 800 mark? Or is there another setting that allows me to OC it more?
Hmm...i know for a fact this RAM goes to 900mhz, i tried loosening the timings to 5-5-5-18 too but same.
Anyway, after trial and error with voltages, i'm upto 3.4ghz with 1.4125v (or 1.39 due to droop). Temps at idle are 2-3C higher across the board, acceptable. Dunno if i wanna try the small FFT's, i know it'll fail at such a harsh test, but gotta be done i guess...
...to be honest, this chip cost me £152, it's not a huge loss as such if it's a dodo
Edit: RAM has hit a wall at 820mhz, so at x8 multi, thats around 3.2ghz, CPU has also hit a wall at 3.2ghz, at 3374mhz now, i get an error on small FFT's, and OCCT crashes.
what..the...hell. This is with 1.425v now. How can both my RAM and my CPU be dodgy, they jsut won't OC well and i've ran out of idea's. How can that ram, which i know can be pushed WELL beyond 900mhz even at 2.1v, not barely pass the 800 mark? Or is there another setting that allows me to OC it more?
Last edited:
I got me OEM version from that place VID 1.325v, I'm happy though now, reached 3.55Ghz rock solid for 3 hours quad prime. Needed 1.488v though (cpu-z reading, not what was set in the bios)
Reaches 73\74c for cores 0\1
Reaches 69c cores 2\3
I tried 3.6Ghz, even tried up to voltage as high as 1.51v....Not happening, one of the cores will fail within 5-6 minutes, tried everything, lower vcore etc....Not a chance, I'm sure i'd need water\me old vapochill to get any higher.
I'm happy with what I've got I think, just about ok for me\what I was expecting.
This is with a scythe infinity running at full 1200rpm (not too loud), in pretty good ventilated coolermaster stacker case with cross flow fan going at 1\3 speed (god that things noisy on high, sounds like wind coming through the windows on a gusty day!)
Dunno whether this is a silly question or not, but i was under the impression DRR2-800 would overclock? Tried 850mhz, basically so i could keep my voltage low by using a x8 multi, 425 FSB, but it doesn't post and saus overclock failed.
Hmm...i know for a fact this RAM goes to 900mhz, i tried loosening the timings to 5-5-5-18 too but same.
Anyway, after trial and error with voltages, i'm upto 3.4ghz with 1.4125v (or 1.39 due to droop). Temps at idle are 2-3C higher across the board, acceptable. Dunno if i wanna try the small FFT's, i know it'll fail at such a harsh test, but gotta be done i guess...
...to be honest, this chip cost me £152, it's not a huge loss as such if it's a dodo
Edit: RAM has hit a wall at 820mhz, so at x8 multi, thats around 3.2ghz, CPU has also hit a wall at 3.2ghz, at 3374mhz now, i get an error on small FFT's, and OCCT crashes.
what..the...hell. This is with 1.425v now. How can both my RAM and my CPU be dodgy, they jsut won't OC well and i've ran out of idea's. How can that ram, which i know can be pushed WELL beyond 900mhz even at 2.1v, not barely pass the 800 mark? Or is there another setting that allows me to OC it more?
I'd set the RAM on "auto", set the voltage to 2.15v (it should automatically loosen the timings as it gets clocked out of spec).
My OCZ stuff needed one of the timings (the one that is the highest, to 42 at 800mhz and 46 at 1000mhz).
The P5K premium only goes up to 46, whereas most other boards go up to at least 20 higher than that, can cause a problem with some memory.
I'm not sure you can actually set it to 46 manually, but if you put all the settings for memory at "auto" it might do it on its own if its programmed into the SPD of the memory to increase it.
Is there a specific forum that Corsair runs they might be able to give you the timings you need\advice on your specific modules. I know OCZ have one and they've been pretty helpful to me
Associate
Those core temps are very high, i read mid to late-60's is max safe zone really.
Wow, that's a huge voltage, i haven't tried past 1.425, which got me just under 3.4ghz, but got a core error in prime95 small FFT's.
OCing the RAM would mean lower voltages and lower temps, as you can use a lower multi. Unfortunately, it won't have that either lol.
Wow, that's a huge voltage, i haven't tried past 1.425, which got me just under 3.4ghz, but got a core error in prime95 small FFT's.
OCing the RAM would mean lower voltages and lower temps, as you can use a lower multi. Unfortunately, it won't have that either lol.
Associate
Ahh thanks hernaldo, didn't think of the obvious. I'll try auto now, soudns like it could work!
'if' it does, i don't see why i can't pass 3.4ghz.
Checked Corsair website, they have my mobo, but no tests of the 2gb PC6400, nearest to it is the 4gb C5 one, didn't help. Had timings real low, vcore 1.3, ram voltage 2.1v, they got 3.3ghz...$%^&ards
'if' it does, i don't see why i can't pass 3.4ghz.
Checked Corsair website, they have my mobo, but no tests of the 2gb PC6400, nearest to it is the 4gb C5 one, didn't help. Had timings real low, vcore 1.3, ram voltage 2.1v, they got 3.3ghz...$%^&ards
Associate
Nope, tried timings on auto, still got the black screen at reboot with nothing displaying and fans staying at that mad startup speed. Then have to hit restart button, it goes on then off then on, then says "overclock failed! ...."
Maybe the memory is just dodgy, though the fact it doesn't even boot with those settings would suggest there's a big problem, though why at 820mhz+...
Maybe the memory is just dodgy, though the fact it doesn't even boot with those settings would suggest there's a big problem, though why at 820mhz+...
Associate
Quick Question: You using Small FFT's test? this makes a big difference to my stability/heat.
On that, my hottest core hits 68C at 3410Mhz.
On that, my hottest core hits 68C at 3410Mhz.
How much voltage are you giving the northbridge? - Try giving it a little more, whatever you're giving it. I don't have a P35 chipset, so dunno what the max they can take is, but if its anything like the X38, up to ~ 1.5v or so should be fine.. obviously keeping an eye on the temps.
That may explain your problems in overclocking the RAM?
Also, the difference between cores still seems very different.. maybe there's just terrible contact between the IHS and the cores underneath? - not much you can do about that really, short of removing the IHS.. but i haven't seen anyone do that with an intel chip.
Oh and try whacking 1.5v through it.. go on!!! (disclaimer.. I accept not responsibility if anything melts)
That may explain your problems in overclocking the RAM?
Also, the difference between cores still seems very different.. maybe there's just terrible contact between the IHS and the cores underneath? - not much you can do about that really, short of removing the IHS.. but i haven't seen anyone do that with an intel chip.
Oh and try whacking 1.5v through it.. go on!!! (disclaimer.. I accept not responsibility if anything melts)
Associate
Ok, i'll try the Northbridge voltage, i think the options are 1.40v, 1.55v, then up, dunno what it is at the moment, on auto.
Well i set relax to 3 in trasnaction booster (disabled) and disabled static write on the memory, so will see if all those 3 help.
As for the vcore, i whacked it upto 1.48v and it took 3.4ghz fine (no extended testing though), yet to see how far that voltage can take me though. Will try again now
Well i set relax to 3 in trasnaction booster (disabled) and disabled static write on the memory, so will see if all those 3 help.
As for the vcore, i whacked it upto 1.48v and it took 3.4ghz fine (no extended testing though), yet to see how far that voltage can take me though. Will try again now