Overclocking the Q6600: Tests with PCMark05

Message Topic Search Topic Options Post Reply Create New Topic Printable Version

   I read Alex's thread about the change in OC policy, and thought I would do a little benchmarking to see what the actual difference in performance would be between various overclock points with the Q6600 Core 2 Quad. Fot those interested, this is what I found, using PCMark 05. The test environment was my Vista Desktop, with all of my usual features and gadgets running, so these scores tend to be about -200 from published numbers on average. Each test was performed under identical conditions.
setting 0: Base score (2.4 Ghz): 6840 PC marks
setting 1: Previous Twisterboost (2.7 Ghz): 7515 PC marks
setting 2: Intermeadiate setting (3.0 Ghz): 8092 PC marks
setting 3: Possible new Twisterboost (3.15Ghz): 8298 PC marks

Setting 1 represents a performance increase of about 9.9% over the stock clock of 2.4Ghz.

Setting 2 comes in at an 18.3% gain over stock.
The performance gain of setting 2 versus setting 1 is a 7.6% increase.

The highest setting I am willing to test on my new rig, 3.15Ghz, offers a 21.31% increase in overall performance.

The performance gain of setting 3 versus setting 2 is a 2.5% increase.

If anyone would like to test at QDR's higher than 1400Mhz, feel free, I can see where this trend is heading and don't see the need to push my brand new CPU that hard.

To me, what this says, is the "sweetspot" for an overclock on the Q6600 is a QDR of 1333. Higher than 3.0 Ghz seems to me at least, to be a case of alot of wear and tear on a CPU for not much to be gained. My impression of the performance changes, aside from the benchmarking, is that there is not that much difference between 2.7Ghz and 3.0 Ghz while actually playing games. So little, in fact, that without an FPS monitor, I would be hard pressed to distinguish between the two settings. There is *no* perceptable difference between 3.0 and 3.15Ghz settings, at least not for me.

For anyone asking for overclock figures beyond that 3.0 Ghz mark for a Q6600, I am going to suggest that the extremely small difference in a synthetic benchmark result, with no realworld gain to speak of, is not worth the extra heat generated, which puts a stress on every other component inside your case.

If anyone would like to offer up additional benchmarks, or thoughts on this, I would be interested in seeing and reading about what others find.

Edited by Tyler Lowe - 29 Oct 2007 at 8:54pm

Author	Message Topic Search Topic Options Post Reply Create New Topic Printable Version
Tyler Lowe Members Profile Find Members Posts Add to Buddy List Newbie Joined: 14 May 2008 Online Status: Offline Posts: 0	Quote Reply Topic: Overclocking the Q6600: Tests with PCMark05 Posted: 29 Oct 2007 at 8:49pm
	I read Alex's thread about the change in OC policy, and thought I would do a little benchmarking to see what the actual difference in performance would be between various overclock points with the Q6600 Core 2 Quad. Fot those interested, this is what I found, using PCMark 05. The test environment was my Vista Desktop, with all of my usual features and gadgets running, so these scores tend to be about -200 from published numbers on average. Each test was performed under identical conditions. setting 0: Base score (2.4 Ghz): 6840 PC marks setting 1: Previous Twisterboost (2.7 Ghz): 7515 PC marks setting 2: Intermeadiate setting (3.0 Ghz): 8092 PC marks setting 3: Possible new Twisterboost (3.15Ghz): 8298 PC marks Setting 1 represents a performance increase of about 9.9% over the stock clock of 2.4Ghz. Setting 2 comes in at an 18.3% gain over stock. The performance gain of setting 2 versus setting 1 is a 7.6% increase. The highest setting I am willing to test on my new rig, 3.15Ghz, offers a 21.31% increase in overall performance. The performance gain of setting 3 versus setting 2 is a 2.5% increase. If anyone would like to test at QDR's higher than 1400Mhz, feel free, I can see where this trend is heading and don't see the need to push my brand new CPU that hard. To me, what this says, is the "sweetspot" for an overclock on the Q6600 is a QDR of 1333. Higher than 3.0 Ghz seems to me at least, to be a case of alot of wear and tear on a CPU for not much to be gained. My impression of the performance changes, aside from the benchmarking, is that there is not that much difference between 2.7Ghz and 3.0 Ghz while actually playing games. So little, in fact, that without an FPS monitor, I would be hard pressed to distinguish between the two settings. There is no perceptable difference between 3.0 and 3.15Ghz settings, at least not for me. For anyone asking for overclock figures beyond that 3.0 Ghz mark for a Q6600, I am going to suggest that the extremely small difference in a synthetic benchmark result, with no realworld gain to speak of, is not worth the extra heat generated, which puts a stress on every other component inside your case. If anyone would like to offer up additional benchmarks, or thoughts on this, I would be interested in seeing and reading about what others find. Edited by Tyler Lowe - 29 Oct 2007 at 8:54pm

sundowner Members Profile Find Members Posts Add to Buddy List Senior Member Joined: 11 Sep 2007 Online Status: Offline Posts: 501	Quote Reply Posted: 29 Oct 2007 at 10:31pm
	Very interesting stuff. Thanks.
	Pro case with extra fan Quad Q6600 2.4 OC'd 3.1! Nvidia 8800GT Asus Maximus Formula Mobo 2xgb 800mhz corsair xfire - xuntiltheendx

Tyler Lowe Members Profile Find Members Posts Add to Buddy List Newbie Joined: 14 May 2008 Online Status: Offline Posts: 0	Quote Reply Posted: 03 Nov 2007 at 8:43pm
	As an update, I've been tweaking all week since I got my system, and I've come to a new level of understanding as to why 2.7Ghz and not higher. I've been able to get the system to boot and remain rock solid stable on 1.2625 Vcore (1.22 Vdrop, 1.21Vdroop), but the instant I go above a 1200QDR I have to raise the FSB voltages which increases heat just enough to take the system from relatively cool running to a bit warm under load. The system will boot into windows at 1.2V FSB, but it won't pass the web rendering test in PCmark without bumping to 1.3V. You wouldn't think a lousy 100mV could make such a difference, but it does. You can really see the difference during stress testing. As an example temp, at 2.7Ghz, my cpu temp as reported by speedfan after the +15C adjustment is around 44C at idle fan at 85% (which I set as minimum since it's quiet at that speed to my ears). Just changing that one voltage level leaves me at 46-47C idle. Still within reason, but 2C makes a difference at full load across all cores. (I know the idles look high, but the ambient is normally at 75-80F. When it was at closer to 65F this morning, idle was closer to 34C) I could probably just have taken DSO at it's word when it comes to OC'ing systems, I mean afterall it's what they do, but I'm the sort of person that has to understand how and why. I'm glad I went to the trouble to learn more though, as it's brought me a whole new level of respect for what DSO does. With all the tweaking I've done, I managed 7789 PCmarks at 2.7Ghz on a Q6600. I'm pretty happy with that, though it won't stop me from trying to figure out if there's a way around bumping my FSB voltage to 1.3 in the pursuit of ~3Ghz. It may not be possible, but the fun with this is in the trying . Edited by Tyler Lowe - 03 Nov 2007 at 8:56pm

donkeypunch Members Profile Find Members Posts Add to Buddy List Senior Member Joined: 21 Sep 2007 Online Status: Offline Posts: 188	Quote Reply Posted: 04 Nov 2007 at 12:49pm
	Thats what I was having fun with trying to understand. From so many of the reviews I've read in articles and forums was making it seem like it was a cake walk to hit 3Ghz on stock voltages. I also had to raise voltages a bit and also wanted to bump up my 800mhz ram. To me it seemed like the majority who had Dominator 1066mhz ram got a 3-3.2 OC while people with 800mhz got 2.7. This of course may have nothing to do with it but was something I noticed that may play a role in the seemingly standard 2.7 OC. I really do like the idea that DS will post different settings so people can try different settings within reason. It should be easier for people at home as well in case they want or need to add anything to help with the OC or control heat such as additional fans, better thermal paste, or whatever is needed to satisfy their OC'ing goals. But again, I think its a great idea that DS will post some different settings for their rigs for their customers to apply and keep them happy. Good times.

Overclocking the Q6600: Tests with PCMark05

About Us

Community

Support

Company

Contact Us

Follow Us