Prescott Benchmark

[ted]

Enough said Here

 

[sman789]

 

[alexrocks]

Or, we could continue not beleiving random foreign sites that don't speak english.

 

[mechBgon]

Give 'em a chance. The early benchmarks for Athlon 64 at XBit Labs didn't look very encouraging but it ended up doing better at release.

 

[Excelsior]

Originally posted by: mechBgon
Give 'em a chance. The early benchmarks for Athlon 64 at XBit Labs didn't look very encouraging but it ended up doing better at release.

Yeah, that and I don't like sites like that.

 

[DivideBYZero]

Originally posted by: alexrocks
Or, we could continue not beleiving random foreign sites that don't speak english.

Just because it's not in english, doesn't make it less true.

Those results however are probably from early silicon and can't really be used to gauge Prescott perfomance.

At least, I hope not, cause they are sheeeeet...

 

[SinfulWeeper]

Well I cant read....
Looks like they cant write either....

 

[Alkaline5]

I don't see why everyone is complaining about the performance here. Clocked at 2.8, Prescott easily outperforms the 2.8C in all but one benchmark and even does better than the 3.2C in about 2.

 

[Ticktanium2038]

Doesnt it strike you funny that pc mark 2002 rated the CPU as slower than the 2.8 Ghz northwood and just edged out Aqunox 2.8Ghz.

But darn it if PCMARK 2003 says its 36% faster it must be.

I hate lying companies. PCMARK!!!

Fully Optimise SSE3 code into a benchmark and claim its real world performance is not a true measurment of performance.

I really am against benchmark companies because they skew reality in favor of who makes the biggest donation.

 

[zephyrprime]

But darn it if PCMARK 2003 says its 36% faster it must be.

I hate lying companies. PCMARK!!!

Where does it say that? I think you meant CPUBench, not PCMark.

I don't see why everyone is complaining about the performance here. Clocked at 2.8, Prescott easily outperforms the 2.8C in all but one benchmark and even does better than the 3.2C in about 2.

The only benchmark where it beat the prescott was the CPUBench 2003 which is a purely synthetic benchmark.

 

[pm]

You can't judge final shipping frequency or performance of a silicon device based on what a pre-release beta part does. About the only thing that you can judge is that the performance or final frequency is not likely to be less than the pre-release part. But even that is uncertain. Pre-silicon samples are used to evaluate a wide variety of things like power delivery systems, thermal dissipation, compatibility, etc. They are not intended to be the final product.

I seem to remember writing practically the same thing when the discussion was regarding the clock frequency of early Hammer samples, so it's not like my reply varies by the product or even the company that I'm talking about. Don't guess at the final product based on anything that a sample does.