cCpuBenchmark Function

Keywords: 	 cCpuBenchmark cCpuSpeed

---

Question:

Have tried to use this function in the CPU extender. The results are mixed. I have struggled to obtain worthwhile results, gathered from a source of 20 PC's or so. Interpretting the data suggests this particular function still has a little way to go before it hits true.

Changing RAM in a PC doesn't seem to effect the return value. I was pleased to learn this through trial and error. However the return value was difficult to believe at times, not corresponding to a known processor Mhz present in the PC. Any ideas on how to make the results more dependable? I haven't played with other parameters, apart from the Numerator ( set to default at 60,000 ). When playing with this figure on my Athlon 750, each time I increase the Numerator I get a considerably different return value. Cheers.

Answer:

I would also try the "cCPUSpeed" function.

The cCpuBenchMark function is merely a couple of tight loops, attempting to calculate how fast a give cpu can blow through a raw loop.

However other factors, such as network traffic and other programs running can affect the results. It is bascially one of the more primitive cpu benchmarks.

The CCpuSpeed function, on the other hand, may actually give you a number that you are looking for.

I assume you are running a current version of the CPU extender?

---

Article ID:   W14527

Filename:   cCpuBenchmark Function.txt