CPU Overclocking

To check the Biostar TForce P965 mainboard at practical overclocking I assembled a system with an Intel Core 2 Duo E6300 CPU cooled with a Zalman CNPS9700 LED, and with Corsair TWIN2X2048-6400C4 memory. I installed a low-speed 80mm fan to cool the chipset heatsinks and memory. The tests were performed in Windows Vista.

Although the memory timings were left at their defaults in the BIOS (4-4-4-12 as was written in the modules’ SPD), the mainboard would set them higher at 5-5-5-12. This is just what you need for overclocking, though, so let’s put up with this defect for now.

Knowing the potential of the CPU, I set its voltage at 1.45V right away. I increased the memory voltage to 2.2V, the FSB Termination Voltage to 1.4V, and the North Bridge voltage to 1.45V. After that I set the FSB frequency at 450MHz. The TForce P965 mainboard conquered it easily. Then I set the FSB frequency at 490MHz which was in fact the maximum the CPU was capable of, and there were no problems again. The system would boot up and pass all tests normally.

The Intel Thermal Analysis Tool refused to work, probably due to the operating system I used, so I checked the mainboard’s stability by running two copies of Prime95 simultaneously. The mainboard passed the test successfully.

I was amazed how simple and easy it was to overclock the CPU on the Biostar TForce P965. If I had ended my tests then, the mainboard would have had high marks only. Its advantages seemed obvious while its drawbacks were few and negligible. Unfortunately for Biostar, I continued my tests with an Intel Core 2 Duo E4300 processor and found a whole lot of problems.

The particular sample of the CPU was capable of overclocking to a 360MHz FSB as I had made sure on an ASUS Commando mainboard, which was based on the Intel P965 chipset, too. The CPU has a higher multiplier (x9) in comparison with the E6300 and a lower default FSB frequency (200MHz). I started my tests from 350MHz FSB, but the mainboard would start up and then freeze immediately. The Self Recovery System, which was supposed to roll the parameters back at over-overclocking, would not work. I somehow managed once to enter the BIOS and lower the frequency to 340MHz, but I had to use the Clear CMOS jumper afterwards.

Starting from 330MHz FSB, the mainboard refused to start up altogether but the Self Recovery System worked fine – the mainboard would restart automatically with default settings after an unsuccessful start. Unfortunately, I had to lower the frequency for very long because the OS booted up at a frequency of 260MHz FSB only. I didn’t perform any stability tests because the result was unsatisfactory regardless of the mainboard’s passing or failing them.

I eventually discovered more problems about the mainboard in its regular operation mode, too. As I mentioned above, the mainboard does not read timings from the memory modules’ SPD. My attempt to select the default 4-4-4-12 was a failure as all diagnostic programs would show 3-4-4-12. Corsair TWIN2X2048-6400C4 is good memory, but I doubt it can work at 800MHz with a CAS Latency of 3.0. So I guess the displayed value was just wrong.

The TForce P965 was not doing quite well with USB devices. My USB keyboard would periodically shut down, just stopping any work. Data could be written onto a USB flash drive, but the system would freeze on an attempt to read it. The access indicator was blinking, but nothing happened. I would have blamed unstable Windows Vista drivers, but system was also unable to boot up from the flash drive with the same symptoms: the mainboard sees the boot device, begins to start up and freezes while the access indicator is blinking all the time.

There were some other troubles. For example, Vista would often find some problems on the hard disk and run Scandisk after the system had been shut down or reset normally. But I’m not sure if it is the mainboard’s fault after all. However, the overall stability of the system based on the Biostar TForce P965 was rather low.