New to the Forum, but hello!

Finished ordering all the parts a couple days ago. Will be assembling them next Tuesday/Wednesday. Goal for this build is a 24/7 stable 333x9 (2.97 GHz) Q6600. I will however also attempt to push it to it's limits on air and test out two TIMs in the process and lapping! I'll also push the limits of the board and memory as best I can (This was "technically" a budget build).

CPU: G0 Q6600 (Batch L737A823) - I checked around and saw a couple people running 24/7 stable at 3.4 GHz @ 1.38 vCore from this batch. This is decently encouraging!

Memory: Transcend AxeRAM DDR2-800 @ 4-4-4-12 1.8vDIMM

MB: DFI Blood Iron P35-T2RL - I'm a bit "concerned" about the passive HS on the NB. I don't think it will be an issue as the Quad shouldn't go too much higher than 400 Bus.

HDs: Seagate Barracudas in RAID 0

VGA: HIS X1950Pro

Case: Cooler Master 690

Case Fans: Scythe Slip Streams (Already had these running and they are pretty quiet for the air they move).

PSU: Cooler Master Real Power Pro 650W

Heat Sink: Thermalright Ultra 120 eXtreme + Cooljag Everflow 110CFM PWM Fan

I will be assembling the rig with Arctic Silver 5 and allowing it to cure. After the curing period I will check Load/Idle Temps by using Core Temp Beta. Unfortunately I'm not the best at making graphs, so I will include the highest, lowest, and average over 1 hour of dual Orthos Small FFTs (Setting CPU Affinity for each one). Idle will be tested by allowing it to sit for 30 minutes following the test. Temps will be converted to Delta T against room Temps as tested by a Fluke HVAC Digital Clamp Meter w/ Temp Probe. The data points I will be looking at are the CPU Clock Speed, Vcore at Full Load (as measured via CPU-Z), Idle Temp (Avg All Cores), Avg/Max Load Temps (Across all Cores), dT versus the ambient air at the case intake (using a Scythe Kama Bay to intake fresh air from the front of the case to the Heat sink, will post pics/videos of the cooling setup), and dT between cores (Hottest Core - Coolest Core).

After that I will overclock to the goal of 333x9 and once again check temps. Then will be finding the maximum overclock allowed while keeping Core Temps below 60 C.

During all of that I will be finding the maximum overclock on the memory up to 2.2 vDIMM at 3, 4, and 5 CAS. Also will try to get the NB up to 425 MHz Bus (Won't be passing this mark with the passive heat sink).

So after ALL of that is said and done... I'll take the Heat Sink off and apply MX-2. I'll allow the MX-2 to "cure" (even though it says it doesn't need to) for 48 hours and repeat the tests (Stock, 2.97 GHz, Max OC). The TIM that is able to maintain the best cooling (as measured by Delta T versus ambient air) will be chosen for the "next" step.

Last step will be lapping the HS and Processor (This will be 1-2 months down the road). Following that I will apply the "winning" TIM and once again repeat the tests.

Prior to loading an OS the BIOS will be flashed to the 9/13/2007 revision and Memtest86v33 will be run from a Boot Disk with the Memory running at stock settings to ensure it's 100% good to go. "Stable" memory overclocks for 3, 4, and 5 CAS will be made at a maximum of 2.2 vDIMM as determined by completion of SuperPi 16M. I won't be going over 2.0 vDIMM or 1.375 vCORE for any 24/7 Operation.

I will also work to include video of the build (specifically about cooling setup) and as a reference for noise. All testing will be done with a closed case as if it was a normal setup.

Thanks for reading, and I hope you enjoy what's to come, I'll try to keep it updated!

Pulled the Processor out of the box and checked it with a razor. It is just slightly concave. Not as bad as some I have seen.. but we'll see how it does with the 120X.

I will be very interested to see how your test goes. I just completed an overclocking test on the Q6600, with a Tuniq Tower 120 here http://www.ocforums.com/showthread.php?t=532466 - I found that core temperature limited the degree of conservative, reliable over clock frequency to 3.5 GHz. Lets see how the Ultra Extreme does.

After reading your results I'm pretty happy I didn't end up buying the Tuniq. I was looking at prices and the Tuniq w/ Fan was about $20 cheaper than the setup I will be installing.

MX-2 did what it said, no curing time. I'm using it and there's no different result from the first time i turn on the rig and the next month i turn on the rig.

Did a quick Benchmark on the "old" system. Have the quad core sitting next to me.. waiting on the shipment to get here Monday/Tuesday.

P4 3.2 GHz (Stock)
2 GB Corsair ValueSelect DDR400
XFX 7600 GT OC'd to 650 Mhz Core/770 MHz Mem

Super Pi 1M 43.532 sec



Lost Planet Extreme Condition DX9

Min Settings 0xAA 0xAF
1280x1024x32

Snow: 28.1 FPS
Cave: 10.7 FPS

Max Settings 4xAA 8xAF
1280x1024x64

Snow: 11.0
Cave: 11.0

Just some quick Benchmarks for fun for me to compare my old system to the new one. Thought it'd be fun to see the difference.

P35-T2RL Blood Iron
2 GB Transcend AxeRam DDR2800 @ 4-4-4-12 1.8v
Q6600 G0 Step

1.28 vCore in Windows

Clock...Ambient...Idle...Load...Max...dTA...dTL
3.0 22C 34C..50C....52C...28C....16C



7 Hours and 15 Minutes Prime Stable, CoreTemp was logged all night.



AS5 has been on it about 6 hours at this point.. so I'd expect temps to drop 2-3C over the next week or so. The MX-2 gets here tomorrow and I'll be waiting on this AS5 to cure up before I abandon it. Temps aren't horrible.

Got my vCore down to 1.264 and stable in Windows @ 3.0 GHz.

I love this chip.

Running Prime on it now. 20 minutes in and the max I've seen is 56 C. Running at 400x8 @ 1.264vCore under Load.

jchunter: Question for you; were your vCores taken at idle from Windows, from the BIOS, or load from Windows?

Edit: nm.. re-read your post and it does say CPU-Z vCore at load from Windows. Looking for stability testing now while the AS5 cures. I'll be doing Benchmarks this weekend for temps.



So far so good! Didn't restart after gaming with it setup like this.. Ran TF2 for 2 hours rock solid. Figured I'd give Prime95 a shot. If it passes 8 Hours of CPU Burning then I'll run a blend on the entire system and hopefully it'll pass. If it does this will be my 24/7 setup. As I said I'll be doing some temp testing this weekend. Computer has been on for a little over 31 Hours now (constantly being restarted to tweak the BIOS, install windows, drivers, etc...). Saturday night will give me about 100 hours on the AS5 so I will bench on Sunday at a few different frequencies. The following weekend I will do a final temp pass through with the AS5 fully cured and replace the TIM with MX-2. Hopefully I can keep the schedule up properly. Before I do benching I also might swap fans.. The 110 CFM Everflow needs to be modded to fit into the clips on the Ultra 120. Currently using a 90 CFM Slip Stream.

Looks good!



Prime Stable 8 hours @ 400x8, vCore 1.264 under load. Temps actually came down over night. Max I saw was 56C and it was hovering on 55C when I went to bed, sat at 54C when I got up. Average over all 4 cores was 53C over 8 hours of Small FFTs.

Ran blend while I was at Factory training all day. Looking good. I'm considering this stable if it makes it all night.

Computer was booted at 400x8 last night around 6 PM my time. Played TF2 on it with no problems until 9:30 PM and started Prime95. Ran Small FFTs until 5:30 AM and then switched to Blend until I got home this afternoon at 3:30 PM.

If it makes it all night... I think it's good.

So I took a table saw to my Everflow fan that wouldn't work with the clips on the Thermalright. It works now.

Did a quick test and it's dropped temps by 2 C on all cores and dropped Chipset temp by 5 C (adds more 30 CFM over the other fan).

Edit: Checked room temp, it is 23.5 C (Via Fluke 902 HVAC Meter with Type-K Thermocouple). Cores never spiked over 52C in 20 minutes. Old setup was pulling 55 C in 20 minutes at 22.3 C ambient temperature.