Slashdot Mirror
Centrino Mobile Equals Desktop Pentium 4 in Speed
Spy Handler writes "On Wednesday during the launch of its new Sonoma Centrino Mobile, Intel put on a demonstration running a video game on a laptop. It matched the performance of a high-end Pentium 4 desktop running the same game, declared Intel. The contenders were a laptop sporting a 2.13 GHz Pentium M processor, 1GB RAM, and the Alviso chipset versus a desktop with a 3.6 GHz Pentium 4 with hyperthreading, 1GB RAM, and the Grantsdale chipset. Is this a testament to how far the Pentium Mobile architecture has come, or a sad comment on the clockspeed-pushing design of the Pentium 4?"
> Is this a testament to how far the Pentium Mobile architecture has come, or a sad comment on the clockspeed-pushing design of the Pentium 4?
No, it's probably only a testament on not showing enough info about the benchmark system/conditions to provide any useful technical data, but only marketing data.
Who know? Maybe the game was simply framerate limited by the similar integrated graphics chipset.
I'm not saying that the Pentium M isn't fast, or as fast as a desktop P4; only that probably that demo don't prove that.
Just my 2c.
Bye!
SeqBox
Billy
Yes. Hell, i would put two, but they're still to damn expensive (and so are the motherboards for it). I hope that changes in the near future though, the P-M is a terrific processor.
...with the fan attached or without?
I had the "pleasure" of performing a heavy number crunching on a P4 laptop. Luckily it was winter and one of the rooms in my house is unheated. Leaving the laptop there (temp. about +3C) with bottom lifted off the floor by some books to allow free access to the built-in fan prevented it from entering thermal throttling mode and allowed it to run at full speed...
45 5F E1 04 22 CA 29 C4 93 3F 95 05 2B 79 2A B2
I 'upgraded' from a P3-1 ghz to a P4-2.26 ghz and noticed hardly any difference. I upgraded from my P4 to an Athlon64 3400+ and it not only smokes it, it also has a variable clock speed which only ramps up when needed.
I've been a loyal intel user since the Pentium 90 came out, but after building several cheap and stable AMD systems for friends and family I took the plunge myself, and I'm more than happy.
Hal Spacejock: Science Fiction with Nuts
Its time to do what we used to do back in 1990 before the Pentium arrived, run benchmarks to determine how fast the machine is.
The only interesting thing about using a game as a benchmark is if the thing will run. Its not unusual to find that a game simply does not run on a laptop.
Looking for an Information Security student project suggestion?
Try http://dotcrimeManifesto.com/
I took the plunge and bought an Athlon 64 3700+ laptop (a Compaq Presario). Not only is it cheap, at well under US$2000, but my kids say it plays HalfLife 2 a lot faster than their P4 3.4ghz system with whatever was the fanciest graphics card a few months ago.
I don't see any reason to roll the dice on Intel again based on the price/performance of the Athlon 64 these days.
Cheers,
The poser of the question (that started this thread) signals his ignorance of microprocessor design and underscores what AMD has said all along, and everyone else who hasfollowed the industry since when there was much more competition in the microprocessor, namely you can't juxtpose microprocessors on clock frequency alone. Anyone remember the Intel i860? Or when MIPS was a stand alone company competing against Intel (seemingly). If you say no to these things, that probably explains why you're even pondering this stuff. Nothing to ponder, some of us have known this all along.
One area where Pentium M is fantastic is in scientific and engineering simulation software. My company produces such a piece of software called OrcaFlex (www.orcina.com). The code is mainly old fashioned 8087 FPU instructions doing 3 dimensional vector operations.
In the past few years clock speed has become much less important than memory architecture in determining how fast the simulations run. Of current architectures P4 stinks and is comprehensively stuffed by Opteron. However, PM even beats Opteron. Our fastest machine for OrcaFlex is a DELL Centrino notebook! This just edges out our top of the range Opteron workstation.
Has anyone else out there seen anything similar with other applications?
It's a valid point. the question: "Is this a testament to how far the Pentium Mobile architecture has come, or a sad comment on the clockspeed-pushing design of the Pentium 4?"... I'd say the latter. The pentium mobile architecture hasn't come a long way, it's been dragged along by AMD's use of similar technology (hell, and even IBM with its PPC970) to run better at lower clock speeds.
It's a sad comment on how damned long the clockspeed-pushing went on for.
You do have a point here: for most games all you need is a decent graphics card; I mean, in some cases you don't even need that: fable on the XBox look very good and is fun to play. However I must say that my IBM T42p (dothan) laptop just feels much more responsive than my P4 2.6 desktop, especially for scientific applications having a 2MB cache makes all the difference!
Would someone explain to me the point of having a wifi device inside the processor? Does it vastly improve performance or reduce power consumption? I personally would think a typical addon card is preferrable. If your network changes usb standards, you don't have to replace your processor.
Exactly. The only news here is Intel essentially admitting their mistake with the marketing driven P4. For those who are surprised by these results see previous stories on the subject. See this Doom3 and Far Cry benchmark from the link in the first slashdot article and this extremetech article and this French benchmark. And these are not the only sources. The fact is that on a modern platform the Pentium M is quite competitive with not only a P4 at nearly twice the clock speed, but also with Athlon64 chips at nearly half the power of even a 90 nm Winchester Athlon64 with a max TDP of either 21 or 29 Watts for the older and newer chips respectively.
That's not to say that it is competitive in every domain, but for gaming it is tough to beat. And, yes, many modern games do scale with CPU power.
Quite an experience to live in fear, isn't it? That's what it is to be a slave.
I had a professor who still swears by the P3-based Xeon for his work and that it will always smoke anything that the P4 has to offer. Why? Strong integer performance.
The professor I speak of is Bob Hyatt, and his research is on computer chess (specifically Crafty, the chess engine we all know and love). The reason the P3-family of chips does such a good job with it is because of the strength of integer calculations. Dr. Hyatt has repeatedly stated that there is not a single floating point instruction to be had in Crafty.
However, the FP unit in the P3 sucks big time. Intel made a processor with a much longer pipeline in order to improve floating point performance--FP is now world class, but the integer stuff won't be as good as it was with the P3 family. (The shorter pipeline is what makes the Athlon a superior performer in some aspects to this day.) This is why we slashdotters are always screaming that raw clock speed will never indicate the supreme chip.
Isn't the Pentium-M based off the PIII core? In that case a PIII that is modern as you put it is the Pentium-M which matches or outperforms the P4.