Intel Core 2 Duo E6300 Overclocking on abit IN9 32X-MAX Wi-Fi

We performed our overclocking experiments on an open testbed that was configured as follows:

• abit IN9 32X-MAX Wi-Fi, v 1.00 mainboard;
• Intel Core 2 Duo E6300 CPU (1866MHz, Allendale B2, 266MHz FSB);
• 2 x 1024MB Corsair Dominator TWIN2X2048-9136C5D;
• NVIDIA GeForce 8800 GTS 320MB graphics card;
• Maxtor 6L200M0 HDD, SATA 200GB;
• Zalman CNPS9700 LED cooler;
• OCZ GameXStream GXS700 PSU (700W);
• Windows XP SP2 OS.

Our test CPU can work at 490MHz FSB (1960MHz in quadrupled values) if the Vcore is raised to 1.45V. We failed to set these parameters from the very beginning, so we decided to dig deeper into the overclocking process to find out the peculiarities of mainboard’s operation at higher frequencies. It proved capable of booting the OS at FSB frequencies up to 1750 (437.5) MHz, then follows the so-called FSB Hole (the frequencies, when the mainboard wouldn’t even start), and then you can boot it again at 1850 (462.5) MHz. unfortunately, the mainboard remained stable only at FSB frequencies of 1600 (400) MHz or less. In this case the system was running stably without any processor Vcore increase.

I would like to specifically stress this fact, because only processor core voltage could be increased freely. It was actually not completely free, as the mainboard started generating hissing sounds when the Vcore went even 0.01V up. It could be one of the coils, although all of them are Ferrite Choke Coils. The sound got louder as the voltage increased. As for increasing the Vchipset parameter, the mainboard was losing stability even at the proven 1600 (400) MHz frequency in this case.

We ran the tests with the very first BIOS version 1.0 dating back to 12/25/2006. I was pretty sure that there were no newer versions, because there was no information on the abit USA website. Later on I discovered version 1.1 from 02/13/2007. It could be the devilish number “13” that played the dramatic role in this case, but the new BIOS fixed only some of the problems and even added new ones to the list. For example, HyperTransport voltage was now set only at the maximum 1.4V, however, we could raise the CPU VTT Voltage and NB Voltage without losing any of the stability and there was no FSB Hole around 1800MHz any more. Unfortunately, these changes didn’t help us get beyond the notorious 1600 (400) MHz FSB during overclocking.

A new mainboard based on a new chipset that hasn’t been studied inside out yet is a great chance for us to perform some comparative testing and reveal the performance peculiarities in different work modes. We were going to run the tests in overclocked mode, because we have already studied the nominal performance of solutions on the same nForce 680i SLI chipset in our previous reviews. Unfortunately, abit IN9 32X-MAX Wi-Fi didn’t give us this opportunity for the reasons described above. But we still performed a few tests with some really astonishing results. Let me explain.