I've been researching the field of overclocking Intel CPUs for several months now and specifically read up on how the TPower I45 works with the E8400 chip. All I was waiting for was some cooler outside temperatures to help keep the family room, where the desktop is located, from becoming an oven due all the increased heat coming out of the case. Cooking my family in the house was not one of my projected benefits from overclocking. :shock: I'll be posting some pictures in my E8400/TPower Overclocking album to show the progress over time. While I'm not aiming for any extreme overclocks, I do hope to get a decent performance increase in the end that doesn't present any threat to the system components. More to come when I've managed to hit some stable settings and can post the pictures to prove them. ;)

8/22/09 Edit: Outside temperatures bottomed out to the low 50s so there's no reason to fear cooking the family this morning with all my stress tests. :lol: Amazingly I never hit any FSB walls, had no Windows boot errors, and every benchmark/component test I've run has passed with no errors. I did run into some heat issues which was expected as I'm aircooling with what is regarded to be a smaller heatsink, the Zalman CNPS9500AT. After one hour of the large FFT torture test on prime95, Core Temp 0.99.5 topped the cores out at 73 and 70, Real Temp 3.00 topped the cores out at 70 and 65 and the TPower onboard CPU temp monitor topped out at 62. While this is a wide range of temperatures all but the highest temp from Core Temp are below the Intel Thermal Specification of 72.4 for this chip. From my own Real Temp 3.0 calibration attempt I figure the real core temps lie closer to what it is reporting; though I'd still rather play it safe and not attempt a higher FSB or a much higher vCore with this cooling setup. Yes I could max the 9500AT PWM fan out on startup, I could max the coolerharbor fan to maximize the pull off the 9500 and I could wait until winter time and use the even colder ambient air to help gain a higher overclock. But the 400 FSB I have achieved for a 3600 MHz overclock is probably safer for my system and more than adequate for my PC's use. You only need so much processing power for word processing, internet surfing and casual gaming. Though the casual gaming can always use some help every now and then. ;) I also planned to call the overclocking quits at 3.6 GHz due to a powersupply restriction. Right now there is plenty of overhead in wattage, but with my recent looks into an ATI 4890 or nVidia GTX 275 GPU upgrade I was amazed at what the full load wattage requirements were. In many models the full load wattage readings topped out at 360w! And here I thought my 700w Thermaltake PSU would be good for several years of GPU revisions. :? So instead of risking the potential to overwork my PSU with a heavy power drawing GPU, I decided to keep my processor overclock to a reasonable level that would leave plenty of wattage overhead for the eventual upgrade. Now I just have to get the money for said upgrade... :oops: I'll post some screenshots of the overclocking process again today as they come in and hopefully some more benchmark results to show the total improvement to the system.

8/23/09 Edit: After 4 hours of torture tests, running several benchmarks and monitoring temperatures I've just about reached the limit as to what software checks I can run on this overclocked setup. Before calling this setup even remotely stable I've been monitoring errors from restarts, cold boots and program crashes. The Vista performance and reliability monitors and logs have been accessed more times in the past few days than during all of last year. :lol: So far 2 out of 6 restarts have produced a reboot error; each one different. Two programs have unexpectedly crashed during use as well. From what I've learned in the forums and TweakGuides is that any crashes, instability or boot problems on an overclocked system should always be attributed to the overclock itself. This is especially true when a non-overclocked system was proven to be stable over a long period of time. If problems persist I know I can make a few changes to the RAM timings and voltage as the current ones were experimental at best. From there reducing the FSB, and the core overclock, would be next on the list. Thankfully the Biostar TPower I45 comes with ten CMOS slots to save BIOS settings for easy saving and reloading of configurations. My settings from the 24/7 stable stock settings are stored and I will not hesitate to bring them back up just for peace of mind and a super stable system. ;) With a few more days of cold boots, restarts and program monitoring I should be able to report on just how stable this current overclock is. I did manage to validate this overclock with CPU-Z and will update it with any subsequent changes to the system.