Thanks I will take a look at that.
Actually when I did the tests (100% CPU) it was very warm in my room and the airflow is not optimum where the case is located. I do run the cooler in a push pull configuration which really helps.
Still it is a mystery why 3.9 seems a hard limit. Lowering the multiplier and raising the bus speed does not help, in fact a higher bus speed seems to contribute more to the increase in temp than a higher multiplier. Also I tried to tighten the memory settings but it seems that is not it either.
I have 1600 memory so I set the bus to 800 (x2), I suppose it is possible to set it to 533 (x2), use better timings and then boost the bus speed in conjunction with the multiplier. but I really do not know if it will make any changes.
I tried up to 2.6 on the HT speed, I assume boosting that will not really help, someone any experience with that? I know if I raise it to 3.2 the system does not even boot but I have not tried anything in between.
What I suspect is that with the Biostar board, which is a nice and affordable board and but certainly not a high end board pushing the system with 6 cores simply reaches some kind of a limit.
I was able to run many stress tests with 3.875 Volts but one time I got the BSOD so I had to raise it to 4.0.
