Overclocking, The Natural Way

The Intel Pentium 4 530 has a stock frequency and voltage of 3.00GHz at 1.3V. Meanwhile, the Abit AW9D motherboard supports a maximum core voltage of 1.7625V and a maximum memory voltage of 2.65V (the OCZ Flex XLC memory officially supports up to 2.40V). Beginning this natural overclocking endeavor, we had no problems reaching 4GHz. With cooling on our side, we were able to reach this gain of nearly 33% in no time -- though not too impressive considering other overclockers have surpassed 6GHz. When the Pentium 4 530 was running at 4000MHz the OCZ Flex XLC had showed no signs of stress. The Ubuntu Feisty Fawn had also successfully booted at these speeds.

However, when attempting to surpass an FSB of 300MHz we had quickly run into a wall. At 4.50GHz, the system was not stable even while supplying 2.45V to the RAM and 1.7V for the CPU. The DRAM frequency was also adjusted but this had not positively affected the overclock. Further tweaking had occurred, but finally when attempting to push the system further with greater voltages the motherboard had finally given up. However, in this experiment we were able to see the OCZ PC2-9200 Flex XLC memory run at its DDR2-1150+ speeds, and in general this was extremely great memory -- though it will set you back over $450 USD.

While this was not a scientific overclocking experiment as it was truly spur of the moment in order to take advantage of Mother Nature, it was definitely enjoyable and we will likely reproduce these tests more accurately once weather permits. The operating temperature when we were running this system had started out below zero but the temperature in the garage had worked its way up to 3~5 Celsius (next time around we are hoping for approximately -10 Celsius). The actual outdoor testing process had lasted for approximately two hours. Even though the Abit AW9D is now in motherboard heaven, our thoughts remain that this is an exceptionally well-engineered motherboard and works reliably with recent GNU/Linux distributions. The OCZ Flex XLC memory had also performed phenomenally through this experiment and we will likely deliver additional benchmarks with this memory in a different environment in the near future.

If you are interested in all of the different frequencies/timings/voltages we had used, our overclocking procedures, or have any other questions feel free to post them in the Phoronix Forums.