Hi
I have the P35-DQ6 DDRII rev1 board and a Q6600 CPU. All the high quality components and impressive CPU PWM design + silent pipe makes it run super cool and ideal for quad core overclocking. So its a great board.
But there is this annoying problem with GA P35-DQ6 when overclocking. Everytime the MCH voltage is set to > +0.2 (0.25 for example) usually after a reboot the PCIEX16 will electronically scale down to PCIEX1 as reported in CPU-Z/Everest and also visible by heavy decrease in graphics performance. I tried F4 / F5 and F6 bios. Restarting the system or shutting down/powering back up doesnt solve the problem.
If you set the MCH voltage to +0.2 or lower it works fine and never falls back to PCIEX1.
It has been confirmed by many GA P35-DQ6 owners that overclock their CPU including me.
Is GIGABYTE aware of this problem? Do they have a plan to fix it?
You can easily recreate the issue I described above.
It works like a switch:
MCH V > +0.2===> PCIEX16 falls back to PCIEX1
MCH V =< +0.2 ===>PCIEX16 works fine.
The problem is you may need more than +0.2V on MCH to have a stable overclock .
To make things clear what I mean by the PCIEX16 slot is the upper most (closest to the CPU) PCIE slot.
Thank you.
I have the P35-DQ6 DDRII rev1 board and a Q6600 CPU. All the high quality components and impressive CPU PWM design + silent pipe makes it run super cool and ideal for quad core overclocking. So its a great board.
But there is this annoying problem with GA P35-DQ6 when overclocking. Everytime the MCH voltage is set to > +0.2 (0.25 for example) usually after a reboot the PCIEX16 will electronically scale down to PCIEX1 as reported in CPU-Z/Everest and also visible by heavy decrease in graphics performance. I tried F4 / F5 and F6 bios. Restarting the system or shutting down/powering back up doesnt solve the problem.
If you set the MCH voltage to +0.2 or lower it works fine and never falls back to PCIEX1.
It has been confirmed by many GA P35-DQ6 owners that overclock their CPU including me.
Is GIGABYTE aware of this problem? Do they have a plan to fix it?
You can easily recreate the issue I described above.
It works like a switch:
MCH V > +0.2===> PCIEX16 falls back to PCIEX1
MCH V =< +0.2 ===>PCIEX16 works fine.
The problem is you may need more than +0.2V on MCH to have a stable overclock .
To make things clear what I mean by the PCIEX16 slot is the upper most (closest to the CPU) PCIE slot.
Thank you.
Comment