Quote Originally Posted by bob511
My experiences:

-ECS P4M800Pro-M v2.0 (mATX), BIOS 061214: displays a warning about the FSB on boot, but allows the user to continue at 266MHz

-Gigabyte GA-965GM-S2 rev1.0 (mATX), BIOS F2, F4: does not even POST.

Still hoping for a solid mATX board that'll take this mod.

EDIT: A more in-depth testing of the ECS P4M800Pro-M v2.0 at different FSB OC settings; this BIOS has a cap of 300MHz FSB in call cases, and "warning" refers to the message about the FSB of a modded CPU at boot (which requires a Y/N choice the first time, and then simply displays the warning afterwards):

Unmodded E4300:
-Disabled: boots at 200 with no warning
-200: doesn't even POST
-201-231: boots at FSB with no warning
-232-265: boots at FSB with warning (at 265 FSB, Sandra showed the AGP bus at 88MHz)
-266: boots at 200 with no warning
-267-300: boots at FSBs between 201 and 225 with no warning (at a setting of 300 and an actual FSB of 225, Sandra showed the AGP bus at 74MHz)

Modded E4300 (insulated BSEL1, traced over BSEL1 tape with run from BSEL2):
-Default: boots at 266 with warning
-200: doesn't even POST
-201-215: boots at FSBs between 268 and 286 with warning, seems tentatively stable at +.03V CPU (effective 1.324 Vcore), unstable at +.02V and below
-216-239: boots at FSBs between 288 and 317 with warning, unstable even at +.03V CPU (probably Vcore-limited?)
-240-265: doesn't even POST
-266-285: boots at FSB with warning, seems tentatively stable at +.03V CPU (effective 1.324V Vcore), unstable at +.02V and below
-286-300: boots at FSB with warning, unstable even with +.03V (probably Vcore-limited?)
(Sandra never displayed an AGP bus clock speed when the E4300 was modded)

Unmodded E6300:
-Disabled: boots at 266 with no warning
-200-265: no option for booting below 266 with an E6300
-266-300: boots at FSB with no warning, seems stable

Modded E6300 (insulated BSEL1, traced over BSEL1 tape with run from VCC):
-Disabled: boots at 200 with warning
-200-265: no option for booting below 266 with an E6300
-266-300: boots at FSBs between 200 and 225 with warning (at a setting of 300 and an actual FSB of 225, Sandra showed the AGP bus at 74MHz)


I tested these all with a 1GB stick of DDRII-667 RAM, set to the lowest memory multiplier in the BIOS. While instability was readily apparent, I make no promises about statements of stability, as I didn't have the time to run Orthos Prime95's Blend for more than a couple of minutes for each of these cases. I'll admit, I don't recall the previous P4M800Pro-M I tried going as high as 300MHz with the E6300, so some of these values should probably be considered YMMV.

Anyway, it seems like the most I can get out of this board and these CPUs is a modded E4300 at 9x285MHz = 2565MHz or so, about a 43% overclock. Sandra couldn't gauge the AGP bus in that case, but I figure that if at 266MHz FSB AGP should be 66MHz, then it should be at a 1/4 divider and it's probably running around 71MHz, which doesn't seem that out-of-spec, so I'm guessing the instability I saw beyond 285MHz is more an issue of the paltry Vcore options...? If that's the case, the manual pad voltage mods might help out; I'll probably give that some effort when I get some time.

For a chintzy ECS board, I guess 2565MHz from a 1.8GHz chip isn't so bad...

EDIT 2: Should also give my week and code--Q641A268--to put the numbers in perspective.

EDIT 3: Tried one of the pad mods to override the CPU voltage--specifically the one VR-Zone pegs as 1.55V, though the BIOS showed it more as 1.5Vish for my particular CPU--and was able to run Orthos Blend for a couple of hours on the modded E4300 at about a 306MHz FSB (set to 228 in the BIOS) x9 = about 2754MHz. I don't know that I'd want to run at 1.5V around-the-clock (it was just the easiest of those mods to tape off quickly), or even if 1.5V is actually needed, but it does at least seem to indicate that Vcore is a factor for my particular CPU. As usual, ECS is typically pretty limited through the BIOS in this regard, so people trying to make the most of the board would probably want to look into forcing a voltage through the pads.

http://www.xtremesystems.org/forums/...58&postcount=8


Check that post