Sandybridge Benchmark vs. my old Q6600 *UPDATED*

[Dribble]

+1, Everyone likes to brag when they hit a high overclock. Rarely do you hear them come back and admit it wasn't stable in the long term.

+2

As another Q6600 owner (which is probably good for 3.3-3.4 but I have run at 3.2 for years) I appreciate this thread. I have my eye on a K series cpu too

 

[Gillbot]

I think it really comes down to:
1) What motherboard?
2) What voltage?
3) Are we only hearing about the successes?

But that's part of my point. You can't compare a "golden clock" when a user has top tier parts. People buy the CPU expecting it to clock 5GHz, but they buy an ECS mobo and fail to understand why they can't reach that clock. There are far too many combinations out there for anyone to reasonably expect a certain overclock.

+1, Everyone likes to brag when they hit a high overclock. Rarely do you hear them come back and admit it wasn't stable in the long term.

exactly, i've seen way too many post up a number then later find out that they had to back it down.

+2

As another Q6600 owner (which is probably good for 3.3-3.4 but I have run at 3.2 for years) I appreciate this thread. I have my eye on a K series cpu too

As above, I went through MANY Q66's and NONE hit that magic 3.6+. I did get a pre-binned one to verify my system could handle those speeds and sure enough, using the same settings the pre-binned chip fired right up at 3.6 and topped at nearly 3.8GHz. When I put the other G0 back in, topped at 3.4GHz no matter what.

I did the same test with a 920 D0. It topped at 3.8GHz, granted it was on a DX58SO which everyone blamed for the low clock. I went through multiple motherboard swaps trying to get that 920 over 4GHz and it NEVER happened. Once again, I got a "proven clocker" and sure enough, the DX58SO fired right up at 4.2GHz without a hiccup.

I too fell into the hype of the G0/D0 and i'm just hoping people won't buy them EXPECTING to hit a certain clock. I just don't feel like locking the inevitable "help me hit 5GHz with my 2600k" thread after thread after thread.

 

[SunnyD]

Overclocking: YMMV. (Notice the period there.)

 

[VirtualLarry]

As above, I went through MANY Q66's and NONE hit that magic 3.6+.

I guess I'm lucky then, my Q6600 G0, the first one I got, did 3.6 on low volts. Granted, it can still hit 90C on a Tuniq, which is no mean feat. These chips run hot!

Then again, I have to give some credit to the mobo it's in, it's a DFI X48 mobo that cost me $260.

 

[Dribble]

I know someone who had a Q6600 @ 3.6. Needed water cooling and 1.5V if I remember correctly. It lived fast and died young (after about a year). Suspect mine might manage 3.6 at 1.5V, but pretty sure it wouldn't still be working today.

 

[Doggiedog]

It's interesting that your bclk is at 99.8Mhz. The bclk on my board defaults to 103mhz. I've got 4.43Ghz at 43x multiplier. I want to get the bclk back down to 100 but for some reason it won't. My system is stable so I can't complain.

 

[Gillbot]

I guess I'm lucky then, my Q6600 G0, the first one I got, did 3.6 on low volts. Granted, it can still hit 90C on a Tuniq, which is no mean feat. These chips run hot!

Then again, I have to give some credit to the mobo it's in, it's a DFI X48 mobo that cost me $260.

I tried that mobo too, no dice. Everyone blamed my "garbage" MSI motherboard which clocked the "golden binned sample" at 3.8GHz, above the DFI's max of 3.6GHz. I suspect that that binned chip would do 4GHz but since it wasn't mine, I didn't want to risk frying it.

I know someone who had a Q6600 @ 3.6. Needed water cooling and 1.5V if I remember correctly. It lived fast and died young (after about a year). Suspect mine might manage 3.6 at 1.5V, but pretty sure it wouldn't still be working today.

None of mine would do 3.6+ at even 1.6v (gasp or more!) with water cooling. Trust me, I threw everything at it I could in the blind frustration of an expected overclock. Repeatedly I was disappointed no matter what I tried. I read every guide, every review, every thread I could find trying every tweak and trick in the book. All were met with simple failure.

 

[webstar]

My Q6600 won't go further than 2.68GHz, how about that? Although, I gotta say I'm not really into overclocking so that's why I'm going with an i5 2500. If by any chance I need an overclock, I'll push it to 3.9GHz for 4 cores with a slight bclk increase and multi at 37. Even with a K version I wouldn't go further than something like 4.2GHz, and for it I would probably have to wait another month or so 'cause all 2500K/2600K cpus were sold out in my country in the matter of days. Guess that retailers didn't expect that.

 

[scrubman]

It's interesting that your bclk is at 99.8Mhz. The bclk on my board defaults to 103mhz. I've got 4.43Ghz at 43x multiplier. I want to get the bclk back down to 100 but for some reason it won't. My system is stable so I can't complain.

Yes it is. I have a feeling its only becasue I'm using the "Auto Overclock" setting but not sure.

I tried that mobo too, no dice. Everyone blamed my "garbage" MSI motherboard which clocked the "golden binned sample" at 3.8GHz, above the DFI's max of 3.6GHz. I suspect that that binned chip would do 4GHz but since it wasn't mine, I didn't want to risk frying it.

None of mine would do 3.6+ at even 1.6v (gasp or more!) with water cooling. Trust me, I threw everything at it I could in the blind frustration of an expected overclock. Repeatedly I was disappointed no matter what I tried. I read every guide, every review, every thread I could find trying every tweak and trick in the book. All were met with simple failure.

It's funny cause I remember starting my Q6600 out at 3.5 and that would be too hot to run Prime95 for 24 hours but I used it for a while. Then down to 3.4, then 2.4, and then back to 3.2 of the 3 years I used it. So if I stick to my plan this time, I will keep my 2600k at 4.4, where I CAN run it 24/7 at 70c max. That may make for a more steady OC life of this new box.

 

[aigomorla]

Go for it! You can probably beat Aigo's overclock too!

never! i will always find a way to 1 up someone on this forum unless someone from XS decides to make my day a living hell.

I did get a pre-binned one to verify my system could handle those speeds and sure enough, using the same settings the pre-binned chip fired right up at 3.6 and topped at nearly 3.8GHz. When I put the other G0 back in, topped at 3.4GHz no matter what.

are u talking about that cherry i sent ya?

I suspect that that binned chip would do 4GHz but since it wasn't mine, I didn't want to risk frying it.

Lulz.. i loved that Q6600 G0.... i it was the best bin'd one i had until i ran into yorkies.

And you guys all remember yorkies. :whiste:


Overclock is always YMMV.
Even if u had my chips, without the cooling, and the rest of the gear, i still couldnt gaurentee you the same overclock.

 

[Gillbot]

are u talking about that cherry i sent ya?

That was one of em. :thumbsup:

 

[jihe]

But that's part of my point. You can't compare a "golden clock" when a user has top tier parts. People buy the CPU expecting it to clock 5GHz, but they buy an ECS mobo and fail to understand why they can't reach that clock. There are far too many combinations out there for anyone to reasonably expect a certain overclock.

exactly, i've seen way too many post up a number then later find out that they had to back it down.


As above, I went through MANY Q66's and NONE hit that magic 3.6+. I did get a pre-binned one to verify my system could handle those speeds and sure enough, using the same settings the pre-binned chip fired right up at 3.6 and topped at nearly 3.8GHz. When I put the other G0 back in, topped at 3.4GHz no matter what.

I did the same test with a 920 D0. It topped at 3.8GHz, granted it was on a DX58SO which everyone blamed for the low clock. I went through multiple motherboard swaps trying to get that 920 over 4GHz and it NEVER happened. Once again, I got a "proven clocker" and sure enough, the DX58SO fired right up at 4.2GHz without a hiccup.

I too fell into the hype of the G0/D0 and i'm just hoping people won't buy them EXPECTING to hit a certain clock. I just don't feel like locking the inevitable "help me hit 5GHz with my 2600k" thread after thread after thread.

I agree, only a few G0 Q6600 can do 3.6 stable.

 

[mlc]

yeah.. it's definitiely tempting seeing some of these 3.6 Q6600 sig's to go back and try to improve my clock...

from what i recall.. i was able to hit 3.4 ok.. but anything above that was shaky..

 

[T101]

I went from a Q6600 at 3.0Ghz to a 2600K at stock. And whow. The difference is like night and day. Encoding video with handbrake as an example: 39 Frames per second average with the Q6600. 239 Frames per second average on the Core i7 2600k.
This is using the same movie being encoded from and to drives that is identical model and make (so read/write speed has little to none effect on the overall performance).
And I was actually not considering upgrading, since my estimates where a 30-40% speed increase, maximum. LOL. Was I wrong....

 

[scrubman]

I went from a Q6600 at 3.0Ghz to a 2600K at stock. And whow. The difference is like night and day. Encoding video with handbrake as an example: 39 Frames per second average with the Q6600. 239 Frames per second average on the Core i7 2600k.
This is using the same movie being encoded from and to drives that is identical model and make (so read/write speed has little to none effect on the overall performance).
And I was actually not considering upgrading, since my estimates where a 30-40% speed increase, maximum. LOL. Was I wrong....

Now that's what I like to hear!! Great info!!

 

[Gillbot]

I went from a Q6600 at 3.0Ghz to a 2600K at stock. And whow. The difference is like night and day. Encoding video with handbrake as an example: 39 Frames per second average with the Q6600. 239 Frames per second average on the Core i7 2600k.
This is using the same movie being encoded from and to drives that is identical model and make (so read/write speed has little to none effect on the overall performance).
And I was actually not considering upgrading, since my estimates where a 30-40% speed increase, maximum. LOL. Was I wrong....

Now that's some of the data I like to see. Not these inflated "bench stable" clocks and unreasonable expectations.

 

[Diogenes2]

I went from a Q6600 at 3.0Ghz to a 2600K at stock. And whow. The difference is like night and day. Encoding video with handbrake as an example: 39 Frames per second average with the Q6600. 239 Frames per second average on the Core i7 2600k.
This is using the same movie being encoded from and to drives that is identical model and make (so read/write speed has little to none effect on the overall performance).
And I was actually not considering upgrading, since my estimates where a 30-40% speed increase, maximum. LOL. Was I wrong....

Is this at stock, or overclocked ?

Either way, the results are impressive..

 

[Modelworks]

I have workstations that use the q6600 cpu overclocked to 3.4Ghz . They are all the same hardware. Gigabyte P35-DS3L, Q6600 G0, Arctic freezer pro heatsink, 4-8GB crucial memory. Never had a problem with any of them and they run 24/7.

One tip is to realize that the bios voltage you set is not the voltage the cpu will run at all the time. That is the MAX voltage it will run at in order to regulate the voltage correctly and allow for the times when cpu usage spikes . For example I set my bios to 1.35v but with a meter I can measure 1.31 at the VRM unless I put the cpu under load then it jumps to 1.35V for a short time and levels off at 1.34V



I really want to go to the core i7 chips but for now it will have to wait till the budget improves.

 

[tweakboy]

Why is the C2Q not overclocked more. Like 3.7Ghz or something. I bet you those scores will be much more closer. No reason to upgrade. show me what your quad does at higher clock speeds then 3.2Ghz,, add about 600Mhz per each core and your scores will be more tight. thx gg and gl, also synthetic benchmarks don't matter. Its real world performance that matters and my DAW is set for many years of music producing. thank you

 

[scrubman]

Say what you will, however I am quite certain that for gaming, with a higher end video card (especially nVidia), a stock SB will give you about 40% - 50% masurable improvement versus an overclocked Q6600. Then, the immeasurable factors like smooth, fluid game play are even greater.

 

[Markfw]

Why is the C2Q not overclocked more. Like 3.7Ghz or something. I bet you those scores will be much more closer. No reason to upgrade. show me what your quad does at higher clock speeds then 3.2Ghz,, add about 600Mhz per each core and your scores will be more tight. thx gg and gl, also synthetic benchmarks don't matter. Its real world performance that matters and my DAW is set for many years of music producing. thank you

Maybe you got your Q6600 to 3.7, but I had 6 total, and not one of them would do 3.7 stable. 3.2-3.4 is the average, and 3.5 was the high of my 6 chips.

 

[happy medium]

Just curious, in games, do you actually experience a huge difference in terms of lag? Does the upgrade make your game play feeling much more fluid?

Did we get an answer to this question? If so I missed it.
thanks.

 

[T101]

Is this at stock, or overclocked ?

Either way, the results are impressive..

That is the 2600k at stock speed. (which is between 1600Mhz and 3800, it turbos a lot).

 

[coreyb]

I bought a G6600 G0 right after they were released. Been running stable at 3.6 since that time. I attribute it to having a good motherboard (asus p5k deluxe), which was a $250 board at the time. I don't know if upgrading would be worth it really....

also, so your average frame rate went up only 10fps in crysis warhead?

 

[scrubman]

Average frame rate really depends on how much of the 1 or 2 minute clip has very complex scenes in it. I think the most important thing is the minimum frame rate which was more than 50% better, providing non stop super smooth and fluid game play. Thats where the difference is really appreciated. The Q6600 is really great and after getting the SSD and GTX570 there wasn't nearly as much "need" to upgrade. I just couldnt stand going more than 3 years without getting a new cpu. And I am very glad that I did. Whoever is perfectly content keeping their old chip should just forget about Sandybridge and move along until the next realease becasue its only gonna get better.