Another newbie overclock

[Teronzhul]

Components first

e6700@2.66ghz
2 Gigs Corsair XMS2 PC6400CR4 DDR2800 @2.1v and 444-12
MSI P6N SLI platinum motherboard
evga 8800gts 640meg A2(currently @ 608 core)
xion 600w psu

c1e disabled
eist disabled
dot disabled
all spectrum settings disabled

currently using stock air cooling, with a zalman 9700 on the way, so I wanted to test some higher settings in preparation for its arrival.

With stock cooling and no voltage bumping I can get to 3.2 ghz using linked settings, fsb@1280 orthos stable for sufficient time to make me happy.

Of course 3.2ghz isn't enough! My ultimate goal was to get to 3.6ghz with a 9x multiplier and 1:1 memory at 400 and an fsb of 1600 with high end air cooling.

Now increasing voltage etc I am able to get to 3.45ghz stable. It is too hot for the stock cooler under load, but boots windows etc fine. vcore around 1.45v.

Moving to 3.5ghz (1400fsb 10x multiplier) I get a ramdump immedietly after boot. Changing timings to 555-18 allowed me to boot successfully, but the system was obviously unhappy. Increasing vcore further seems to have no effect.

Any higher FSB entries and the system won't even fire up to bios and requires a cmos reset.

This makes me think the ram is the weak link? Shouldn't it be good much higher than 350mhz for the ram I'm using at 2.1v? I feel like I have to be missing something. I've done all the research I can without wading through 90 pages of the MSIp6n thread and I'm stumped. Any help would be very appreciated.

Thanks

 

[BillParrish]

if your memeory divider is set to 1:1 your 10x350 should not be memory limited.
Either thats it for the CPU or you need the zalman.

Your I got that, but gonna do this, then did that, want this. is confusing
So I am not sure exactly what you have tried in regards to FSB.

I would suggest while you are waiting for the Zalman that you :

set you memory to SPD , memory ratio to 1:1 (linked) set you cpu mulitplier to its lowest setting, FSB at 320 (becasue you know the board can do that) and see how high you can get the FSB at the lowest cpu mulitplier so you know the limits of the board. Once you know that, and you know the memory should not even sweat at actual memory speeds 800MHz and lower, you can jump up to the next mulitplier start again at 320 or 350 and see if it acts the same (ie. you can get at least to 400 FSB) and then rinse and repeat. Take notes the whole time.

I suggest a systematic methodology to determine the limits of your board and memory then cpu.

I also suggest booting into memtest86+ at a low cpu mulitplier and test the memory for at least one full memtest cycle once you get around 400 FSB and higher so you can be sure you memory works as you expect, first on SPD loose timings then on factory spec timings. I have the same memory (might not have the same IC memory chips) and its good at stock timings to 835MHz and can go a lot higher on loose timings. .

All I am saying is that before you go for max cpu, you should already have an excellent idea of what the boards Max FSB is and validate your memory at or near that Max FSB.

 

[Teronzhul]

Thank you for deciphering my disjointed jumble of thoughts. I'm afraid my writing was a reflection of my haphazard approach to overclocking. I've taken your advice and dropped the cpu multiplier to 6 and attempted slowly stepping up. I placed my memory timings on auto, and increased the motherboard voltages to their highest settings.

Even then, the best I've been able to push the system was to a FSB of 1500mhz (375 memory). Interestingly it seemed stable like this, but as soon as I tried to move to 1520, the system would no longer even attempt to boot. I stared at only a black screen.

My next step is to run memtest at 375 and see what my results are, but I haven't had the chance just yet.

I know that not all systems are created equal, I've seen quite a few screenshots around that boast fsb speeds exceeding1700mhz, and am confused and frustrated as to why I can't break 1500.

 

[Teronzhul]

Well, I guess I really should have tried to trudge through all of the p6n threads. The motherboard seems to have been the culprit, and after reading some other individuals findings I decided to skip 1500-1600 fsb altogether and try 1700 (425mhz).

While my computer won't even try to boot in those ranges, it is quite happy zipping along at 1700 as far as I can tell. How strange.