1.7GHz Pentium M outperforms 1.66GHz Core Duo

In the XPS M1210 review, under the PCMark 04 benchmarks, the 1.7GHz Pentium M PortableOne scored 3274 whereas the 1.66GHz Core Duo Toshiba scored a mediocre 2653.

Is this an error or is the 1.7GHz Pentium M a faster chip than the 1.66GHz Core Duo?

 

For single threaded applications, a Pentium M at a faster clock speed would out perform a Core Duo at a slower clock.

 

Yeah like budding said, the new chips are aimed at multi-tasking, since thats what most people do now. Multi-tasking multiple demanding tasks is what newer chips are very very good at.

 

Also, it would appear you are comparing the M1210's PCMark05 score to the P1 SR2's PCMark04 score, which doesn't yield a true comparison. You'll see the M1210 bested 6,000 in PCMark04, well ahead of the SR2.

 

I thought the Core Duo 1.67 was faster than the 1.7 PM because of the faster FSB?

In other benchmarks im sure i've seen a T2300 beat a 1.7 PM... if only i could find the link...*scratches head*

 

Im afraid i dont think your right. There are several super pi results that prove that infact the core duo is faster per clockspeed than the old pentium M even in single threaded apps.

 

SuperPI is probably the worst benchmark for cpu performance which shows the core duo is significantly faster (60-70% in many cases for single threaded applications) than the equivalent clockspeed pentium-m which is untrue.

However the core duo is faster than the pentium-m even on single threaded applications by about 10-30%. And in multithreaded applications this rose to a massive 60%...

Anandtech is probably the most informative site for technical information. There is a real world comparison of pentium-m vs core duo and they even clocked both to be same.
http://www.anandtech.com/showdoc.aspx?i=2663&p=7





MULTITHREADED APPS:


So in short the 1.66 duo will be still quite a bit faster.

 

There are other applications like Photoshop where the Core Duo is significantly faster as well.

 

So therefore it is safe to say that the PCMark 04 bencmarks was bogus?

I simply cannot come to terms with how a 1.7GHz Pentium M can outperform a 1.66GHz Core Duo.

 

Also notice the difference in performance in identical Turion 64 processors:

Acer Aspire 5002 WLMi (AMD Turion 64 ML-30 1.6GHz) 2,392 PCmarks
Acer TravelMate 4400 (AMD Turion ML-30, 1.6GHz) 3,104 PCMarks

Something is not right here.

 

Perhaps it's a typo. Because a score of 2392 by a ML 30 on PCMark05 sounds about right. It's too low for PCMark04.

 

Yeah...I figured so.

Does the amount of RAM in a system affect the PCMark score? This could be the reason.

 

The Core Duo is faster than a Pentium M, no doubt. The lowest Core Duo will beat the fastest Pentium M in any sort of multitasking scenario or multi-threaded app. There is no reason to get a Pentium M these days; Core Duo or bust.

PCMark doesn't only measure processor, so it is inaccurate to say that the Pentium M is faster based on that (which it is not).

And, PCMark04 has always been an iffy and unreliable benchmark. I never use it. Even when it was the only thing out there, people didn't like it. Take all results from that benchmark with a grain of salt.

 

Thanks Chaz, your comment was indeed helpful in trying to solve this problem.

 

it may win a benchmark, but in the real world Im sure the core duo 1.6 is quite faster. I have benchmarked my averatec and people have beaten my score yet they dont game or multitask as well. benchmarks are great for initial comparisons, but Ill take real world performance for sure.

 

What have you used to benchmark?

 

Chaz what would you recomand as a benchmark software?