Mafna

The voltage is not only in bios but during booting windows up. That is a CPU intensive process so the proc might crap out before booting up windows if it does not get the juice while it needs it to power up the windows while being OCed to levels where it needs more than 1.6v for instance. By the way. Does your Mobo also increase CPU voltage when the CPU is being stressed?

But how does one get the machine to boot up overclocked at levels where a higher voltage is needed, if the processor only gets 1.55v max? I mean, if at the level I want it at it needs 1.6v or more? How to boot it up then? If we put asside overclocking from Windows. I mean, does this limitation in booting up mean it is not possible or is it posssible?

Does anyone else get the feeling that on NF3 at least, the voltage during bootup is only the voltage set by actual number without the +% part? Could this also be the reason for many cold boot problems?

I should have a sig now. Basically if I understand one of the threads here it is even better to use deviders and not overclock ram and instead go for highest possible CPU MHZ for highest performance in real life. Is that correct?

Finally! Thanks. At least I know. I thought it was very weird. Because for this RAM even 255 is a whole lot! I have another question-which bios or what do I do in order to make divider work for me at multi 10? So I could run the CPU at it's max and ram near or at its max? I imagine I should use 9:10 divider and use FSB of 275 or something like that but it doesn't work well. Any suggestions?

Noone knows howcome that RAM works at 320 Mhz at multi 4 but only up to 255 mhz at any other multiplier? Any ideas?

Don't think so. Because the computer works fine at much higher speeds at much higher voltages and in general, the RAM works flawlessly at 255 at any other multiplier but not higher. Only on multiplier 4 does it work at 320 mhz. Flawlessly. At 2.5v. At multiplier 5 it is already hitting the ceiling at 255 mhz. I always have ram set manually at 1:1, CPC ON etc.. If I put it in 2T it makes no difference in terms of reaching higher speeds. Given that the computer works normally at 10*255 at 1.5v it should also work for instance at 275*9 at same voltage or even higher or lower but it doesn't. It doesn't post. Dividers also don't work well and the thing won't post either...

I posted the original message in the wrong department. Could anyone explain this to me? http://www.dfi-street.com/forum/showthread.php?t=18592

For the fun of it I tried what this RAM can do at 200 FSB and it can do it with 1.5,2,2,? timings. ? means any number. I actually set it to 0 as strange as that sounds since it gets most memory bandwidth that way. Please see attachment or link: http://www.shrani.si/pics/ramna155786.gif But to get to back to the main question-what do you think is the narrow throat to what, why, how, ...?

DFI Lanparty NF3-latest original bios Newcastle 3000+ 2*256 Mb Infineon DDR 400 3,3,3 (likes low voltages) Leadtek 6800 ATA Hitachi 7k250 with 160 gigs NEC 3520 LC Power 550w Air cooled but temps not a problem here. I have WIN XP with SP2.

Hi, I usually have my Newcastle 3000+ at 2550 Mhz at 255*10 at 1.5v with RAM at 255 T1 at 2.5v. The RAM won't go further without failing tests. Even with slightly lower multipliers the same ceiling is hit. Now when I play around with multiplier 4 I can get RAM to 318 Mhzat 1T, Vcore 1v and RAM at 2.5v. My RAM likes low voltages. Please see the picture in the attachment or a link at: http://www.shrani.si/pics/rezultat55314.gif So I am wondering, how is this possible if the controller is supposed to be the bottleneck-especially at low Vcore voltages? And it actually also works also with only 1v on the Vcore. Howcome I cannot get even near such FSBs on multi 5 or higher no matter which voltage I use,.. I mean literally-what is stoping what in the machine? Chipset? RAM? Processor?...??