How much does your amount of RAM affect a CPU only test?

[ditchmagnet]

Does it a affect a CPU only score? I had my 920 overclocked to 4.5 with HT, and scored about 9k on a CPU only, then again at 4.7 with HT, and about the same score. I am wondering if the fact that I only have 2GB has created some kind of wall? I know someone else who overclocked to only 4.2GHz, and scored over 9400 on CPU only. The only thing I can think of is the RAM difference.

Unless I am doing something very wrong, it doesn't make sense to me that 4.2GHz would score so much higher than 4.7GHz.

Thanks.

[wonderwrench]

Does it a affect a CPU only score? I had my 920 overclocked to 4.5 with HT, and scored about 9k on a CPU only, then again at 4.7 with HT, and about the same score. I am wondering if the fact that I only have 2GB has created some kind of wall? I know someone else who overclocked to only 4.2GHz, and scored over 9400 on CPU only. The only thing I can think of is the RAM difference.

Unless I am doing something very wrong, it doesn't make sense to me that 4.2GHz would score so much higher than 4.7GHz.

Thanks.

If the physical cores are loaded 100% HT can do nothing. That's why you see no improvement. CPU tests normally run in cache as far as I know so the quantity of ram should make little little if any difference.

[David (PassMark)]

You should notice almost no difference between having 2GB and 4GB of RAM in the CPU tests.

CPU might also be throttling down if it overheats.

[ditchmagnet]

If the physical cores are loaded 100% HT can do nothing. That's why you see no improvement. CPU tests normally run in cache as far as I know so the quantity of ram should make little little if any difference.

HT makes a huge difference on the CPU benchmark, usually gets and extra 1000 points.

You should notice almost no difference between having 2GB and 4GB of RAM in the CPU tests.

CPU might also be throttling down if it overheats.

well, at 4.7 with HT, it might have throttled since it did get really hot, but at 4.5 with HT, it didn't get hot enough to throttle. so what could be causing the huge point gap between 4.2 and 4.5-4.7 if its not the RAM? would the different motherboard give it that much of an increase? I have a GA-EX58-UD5 which is a pretty decent board. The other board is a P6T deluxe. I just wouldn't think that would make such a difference, especially considering the the speed of the CPUs.

Just trying to understand whats going on here.

[David (PassMark)]

32bit vs 64bit is a factor that can have a big impact.

[ditchmagnet]

both tests were on 64 bit OS, and 64 bit version of the program

[David (PassMark)]

Hard to comment in any more detail without seeing the actual results and having the full specs for both machines. Do you have the baseline numbers for these results (if you uploaded baselines).