Tom's Hardware tests first Banias/Centrino/Pentium-M notebook

[NFS4]

http://www.tomshardware.com/mobile/20030205/index.html

 

[kuk]

No comments yet? Damn ...

Well, as always, I found this THG article lacking and confusing. Some tests the P4 appears, while in others, there's no mention.
If AT reviews the Centrino, I would like to see benchmarks comparing it to the PowerMacs, besides the P4 (battery time benchmarks that is ... including DivX playback).

 

[Justorq]

I find the new processor pretty amazing ... especially with 1MB of Cache...

Justin

 

[NFS4]

The performance and battery life look very promising.

 

[human2k]

that Pentium-M 1.6 gave the 2.2p4-m a spanking.....bad boy!

 

[NFS4]

Originally posted by: human2k
that Pentium-M 1.6 gave the 2.2p4-m a spanking.....bad boy!

It'll be funny to see how Intel describes the clock speed differential between the P-M and the P4-M considering that they were preaching clock speed over the Athlon XP

 

[OldSpooky]

Looks like the performance is generally somewhere between a 2.2 and 2.4 GHz Pentium 4. Impressive.

 

[Wingznut]

Before you look at all the pretty graphs and compare the two cpu's (particularly the gaming benchmarks), take note that the two laptops were not configured equally. (i.e. The P4m had a 32mb Radeon 7500, wheras the Pentuim-M had a 64mb GF4.)

But yes... The Centrino is an awesome mobile platform.

 

[bigshooter]

I wish they would hurry up and release these so that I can order my new notebook. I'll probalby havev to wait til summer though . Too bad Tom's had to do the review. Anyone else notice that their reviews become more and more fuddled every time you read one? Anand always has clear and concise reviews, where Tom's seems like it just got translated (poorly) from German. At least when Tom wrote them himself they were pretty good. His reviewers of the week are really not that impressive.

Did anyone notice that they stated that the centrino family would replace the Pentium 4-M? I remember the story about the the mobile P4 being different than the p4-m so that they could ramp higher clock speeds sooner, but are they getting rid of the mobile version with clock speed throttling? I always thought they would keep p4 around for performance notebooks and then aim banias for the business, low form factor machines.

 

[Justorq]

what's the point in keeping the P4-M when the Banias is just as fast if not faster and also costs less and less heat !!!!
I think it's good that they dump the P4-M...

Justin

 

[chizow]

Originally posted by: Justorq
what's the point in keeping the P4-M when the Banias is just as fast if not faster and also costs less and less heat !!!!
I think it's good that they dump the P4-M...

Justin

OC'ers like em bc of the low multi.

Chiz

 

[Wingznut]

I don't believe the P4-m's will be "replaced"... At least not for the next year or so.

We'll have to wait for more (read: BETTER) reviews to come out, but I think you'll see that the P4 is still a more powerful cpu.

 

[Remedy]

It'll be funny to see how Intel describes the clock speed differential between the P-M and the P4-M considering that they were preaching clock speed over the Athlon XP

Centrino 2200XP?

 

[dullard]

It'll be funny to see how Intel describes the clock speed differential between the P-M and the P4-M considering that they were preaching clock speed over the Athlon XP

I hear this argument over and over about Centrino. But could someone, anyone, just show me where Intel said clock speed is all that mattered, and not performance? Don't show me Intel fanboys but show me where Intel said it.

 

[Wingznut]

Yeah, I'm also curious about this "preaching clock speed over the Athlon".

And actually, I can think of a couple times that Paul Otellini (Intel COO) stated that mhz wasn't everything.

 

[Remedy]

Dullard&Wingz. I don't think that anyone is accusing intel of making that comment per se. But what most of these enthuisast are looking at is the high frequecy of the Pentium 4 number wise and looking at the a basic 1.7ghz athlon for example in certain and most benchmarks and are asking; "Well how come the athlon is clocked so much lower but yet, it's keeping pace with one of the former flagships from Intel?"

Now, how do you explain to someone quoting from the above and the "early benchmarks" performance of the centrino compared to the Pentium4m? It's mind boggling to most consumers to even think about.

So, what are the extra mhz of the P4/M being used for?

Sincerely yours,

Joe ignorant.

 

[dullard]

Originally posted by: ReMeDy{WcS}
Dullard&Wingz. I don't think that anyone is accusing intel of making that comment per se. But what most of these enthuisast are looking at is the high frequecy of the Pentium 4 number wise and looking at the a basic 1.7ghz athlon for example in certain and most benchmarks and are asking; "Well how come the athlon is clocked so much lower but yet, it's keeping pace with one of the former flagships from Intel?"

Now, how do you explain to someone quoting from the above and the "early benchmarks" performance of the centrino compared to the Pentium4m? It's mind boggling to most consumers to even think about.

So, what are the extra mhz of the P4/M being used for?

Sincerely yours,

Joe ignorant.

The thing is that processors have always had different IPCs. The 2.0 GHz Celeron is a ton slower than the 2.0 GHz P4. The 1.0 GHz P3 is a ton slower than the 1.0 GHz Itanium. The Northwood is faster than the equivalent frequency Williamette. Prescott will be faster than the equivalent speed Northwood.

This isn't just an Intel thing. The duron is slower than the equivalent frequency Athlon. A 1.0 GHz Mac is faster than a 1.0 GHz P3 or 1.0 GHz Athlon, etc. Don't get me started on the C3 chips.

Any consumer who asks the question: "Why is this Celeron so much cheaper than the P3/P4?" will be given the correct answer. If you are a blind consumer and you don't ask any questions at all then yes you might be buying the wrong chip for your needs - is that Intel's fault? And unless all chips from all the companies all have the exact same IPC, that will always happen.

But the fact remains that Intel has had for years several lines of chips that have the same frequency but drastically different performances. Why would one more chip be any different or any more confusing?

I don't know if there will ever be a fix for this. The only fix I can think of is a thorough set of benchmarks by an independant party giving all chips being sold several performance ratings (based on currently used popular programs). There are some good examples of this (SPEC for science calculations for example). I'd love to see a SPEC score, a GAMING score, OFFICE score, and a VIDEO score right there next to the frequency on every chip made by every company. Sure there will be an odd program out there that goes against the trend, but you cannot test everything. Sure there is room for error, but as long as it is independantly tested you won't see things like Intel only using Quake for the gaming score or Apple only using Photoshop. It won't happen though.

 

[Markfw]

Ditto what dullard said ! (and basically ReMeDy{WcS} also) We really need a standard. The only good comparison that works is the "brake horsepower at the rear wheels" which still has some discrepancies. It is even worse in the CPU community since work, games, scientific, CAD, video and sound editing all have unique requirements. At least there is some compitition nowadays instead of Intel being "the only kid on the block".

 

[Rand]

Originally posted by: chizowOC'ers like em bc of the low multi.

Chiz

One seldom see's overclocked notebooks however, so I'm not even remtotely confident that will have any impact on their relative worthiness.

I'd agree with Wingznut that I don't see the P4-M being replaced, at least for the next yr or so they high end P4-M should still yield a performance advantage. For larger/heavier performance oriented laptops the power savings from the Banias won't matter much relative to the top end P4-M's performance.
Similarly price advantages alone should keep the Mobile Cereron alive for awhile.

Differences in system configuration effectively ensure that gaming performanceresults are of little benefit.
Battery tests are even less useful, as the screens alone are very different and likely have a tremendous impact on battery life let alone the rest of the laptop.
Office application performance should give us some indication of performance however. POV-RAY is of definite benefit.
SiSoft results mean precious little one way or another besides system memory benchmarks which are of mild benefit. PCMark is also effectively useless so far as I'm concerned.

All told, in the benchmarks of merit the Banias performed quite well IMHO.
A broader benchmark suite and more tests similar to POV-RAY or SysMark's Office Application scores would have been nice though.

 

[Resh]

What kept running around in my head while I was reading the review was:

"So Intel has a chip that uses next to no power, a higher IPC than the P4, and will move to 0.09 micron process soon. Fine, sounds cool, but why aren't making a desktop version so that we wouldn't have to suffer with fan noise and heat?"

I appreciate that desktops are all about speed, but couldn't we reverse the normal cycle here? Normally, we take fast, hot desktop processor and slow it down to put in a notebook. Seems to me, we could speed up the Banias and end up with a desktop processor that would be as fast as what we're used to, but that would be cooler and quieter to run.