Intel Quad Core Testing Data

Q6600 Testing Data

Date Log; August 15, 2007

Ambient Temperature; 20 degrees

We had intended to use the Live testing pages to record our overclocking adventures with the Intel Quad 6600 CPU. The problem with this is one of obtaining correct information from Speedfan. The program does not want to play nice with the onboard temperature sensors on the Quad 6600 CPU for some reason. We understand that this is a well known problem and is being addressed by the author of the program. Just to clarify the problem further the Speedfan program subtracts 15 degrees or thereabouts from the total for each core reading.

In a drop back position we will publish Core temp graphics of our data until the problem is resolved with Speedfan and we can again produce live results.

Testing

On to testing. We used our Torture Rack configuration for this CPU. Details of this project with some of the results for core 2 duo dual core can be found here. I better point out that during our testing we are liquid cooling the CPU,GPU,Chipset and hard drive in each series of testing so slightly better results would be obtained with CPU only cooling.

Okay so why cool all of the parts?

Lets face it with new games due out very soon and/or a busy schedule of graphics rendering it seems likely that CPU only cooling would be unlikely. We should add here that we were also inquisitive to find out what a high end gaming machine might be like. Oh! and because we could.

The prior research of a variety of forums indicted that the most common upper range for this CPU version was in the 3-3.2Ghz range. One source used was the overclocking range found here on these OCAU forum pages where approximately 34% of responses were in this range.

http://forums.overclockers.com.au/showthread.php?t=518286

It seems only reasonable therefore that we start our overclocking at the upper level of 3.2Ghz and then if possible move on from there.

Click here for larger Image

Next its on to 3.4Ghz where only 11% of the persons polled (at the time of writing) achieved success. We ran our first "single core" Super Pi test run at this overclock.

Click for Larger Image

For the next section of testing we have updated the Core temp program to version 0.95 to accurately predict VID values. A picture of the new values below.

Date Log; August 16, 2007

Today we are heading for 3.5Ghz.

Click for Larger Image

Then onto 3.6Ghz where we will conduct multiple runs of 32M Super Pi. Just for general interest sake 12% of the persons poled in the survey mentioned above were able to reach this figure. After this point the poled results show a marked decline in numbers reaching 3.8Ghz or higher.

Click for Larger Image

So with rough overclocking carried out its now down to a few weeks of tuning up the CPU. Not a bad result for within stock voltages (1.10-1.372Volts).

Date Log; August 17, 2007

We decided to test the upper limits on the current CPU/ram combination we are using before going on to final tweaking. The upper limits reached using this ram was.

I think it would be reasonable to assume that 3.8Ghz would be attainable with better quality ram and attention to timing adjustments.

Next.....Final tweaking to see if we have anything left in this CPU......stay tuned