Pentium 4 vs. Pentium M

[amphibious]

If I'm ALWAYS going to have my laptop plugged in and the price and specs of the laptops are similar am I better off going with Pentium 4?

 

[dnuggett]

What speeds are you comparing here? And what will you be doing with the notebook/laptop?

 

[GnomeCop]

by what little you posted so far, I would recommend a pentium 4 cpu.
For a more accurrate recommendation, we would need to know what you plan on doing with this laptop

 

[Bateluer]

I'd still take the Pentium M over the Mobile P4s or the P4M's. A PentiumM 1.7Ghz is a match for a 3.06Ghz Mobile P4 in most cases. Even if battery life is not a concern now, but its nice to have.


Edit - Is Intel even going to continue downgrading their high end desktop chips for use as mobile chips, now that they have the Pentium M and the upcoming Dothan chips?

 

[bunnyfubbles]

Probably, there's a reason why they actually sell P4s as their desktop chips if the PMs are actually as good as you make them sound to be. Sure a 1.7 PM might beable to give a 3.0 P4 a run for its money in a few apps but in others the P4 3.0 will utterly crush the PM. PMs are meant for laptops due to their high performance despite requiring/putting out low voltage/low heat to get that performance. PMs are clearly the way to go for low weight, low heat, and unplugged perfromance, mobile P4s are a joke and so are celerons (all P4 based celerons in fact), they may save some power over the flat out desktop P4s but the PM is clearly the answer there. You'll get superior performance from a desktop P4 in a desktop replacement notebook but it will most likely be a little bigger and heavier, but performance hands down superior. If you never unplug (except to quickly and conveniently relocate the rig) then a P4 would be best. Otherwise a PM can be very appealing as it was designed almost exclusively for notebook use.

 

[dnuggett]

Sure a 1.7 PM might beable to give a 3.0 P4 a run for its money in a few apps but in others the P4 3.0 will utterly crush the PM.

Ok, so what apps will it get "crushed"? I'm not calling you out just yet but I'd like to see if this is an opinion w/o fact.

 

[alexruiz]

Originally posted by: dnuggett

Sure a 1.7 PM might beable to give a 3.0 P4 a run for its money in a few apps but in others the P4 3.0 will utterly crush the PM.

Ok, so what apps will it get "crushed"? I'm not calling you out just yet but I'd like to see if this is an opinion w/o fact.

Sysmark 2002 is a very good start here.... the IPC of the P4-M is higher than the P-M according to this benchmark....
Anything that is memory intensive (P-M FSB 400 MHz) will show similar results. The true strength of the P-M performance resides in the 1 MB L2 cache....

I don't know about media encoding, but maybe we can get help from the guys who own similar machines......

In response to the original post, go Athlon 64! Your usage pattern fits perfectly for a nice 64 DTR laptop.

 

[dnuggett]

Sysmark 2002 is a very good start here.... the IPC of the P4-M is higher than the P-M according to this benchmark....

Ok that is synthetic. He said apps. I'm looking for the apps that get "crushed". I'm not looking for a post naming memory intensive apps either, unless it can be shown specifically and not theoritically.

 

[alexruiz]

Originally posted by: dnuggett

Sysmark 2002 is a very good start here.... the IPC of the P4-M is higher than the P-M according to this benchmark....

Ok that is synthetic. He said apps. I'm looking for the apps that get "crushed". I'm not looking for a post naming memory intensive apps either, unless it can be shown specifically and not theoritically.

According to babpco, it a summary of used applications....

You are asking for something that has not been done before, as the usual set of applications used to test desktops have not been used in laptops. I was looking for reviews of laptops, and none of them tests anything beyond winstone, sysmark and battery life.

I own neither of them, so I guess we can ask for the help of other owners to run some set of specific benchmarks. However, let's not lose the objectivity.

Let me see if I can borrow my brother's laptop to serve as sparring (Uniwill N251C2, barton 2500+, kt333, mobility radeon 9000) I think where the P4 beats the Athlon XP it will surely beat the P-M, and viceversa.

Could you prove mw wrong?

 

[manko]

I'm guessing the P4 beats out the P-M on MPEG-2 encoding (as it does the AMD chips), which could be an issue if you use your machine to edit DV video and want to output to DVD, for example. Unfortunately, I haven't found any reviews that test media encoding with a common frame of reference making it difficult to compare the P4 and P-M. All reviews I've found use different encoders with different settings and measure on different scales.

 

[alexruiz]

Originally posted by: manko
I'm guessing the P4 beats out the P-M on MPEG-2 encoding (as it does the AMD chips), which could be an issue if you use your machine to edit DV video and want to output to DVD, for example. Unfortunately, I haven't found any reviews that test media encoding with a common frame of reference making it difficult to compare the P4 and P-M. All reviews I've found use different encoders with different settings and measure on different scales.

You are exactly right! There are no reviews of the P-M running the test suite used to test the P4 vs Athlons.... I guess we need to start creating the data ourselves.....

 

[Bateluer]

Who's got a 1.7Ghz PentiumM that they can bench against a 3Ghz P4?

 

[manko]

Here's an example of what I'm talking about, from that P4 vs. Athlon 64 article: Attack of the Mega-Notebooks: Test Results: Media Encoding .

If you take a look at those charts, they are only useful to compare those two machines, otherwise there is no frame of reference to compare them to machines in other reviews. They test Quicktime MPEG-4, Windows Media and Divx 5.1, but they give no indication of the original video clip's resolution, length, frame-rate or anything of use to use their results to compare to tests of other machines on other sites. Their numbers tell us next to nothing.

The most useful encoding test I have seen anywhere is in a Japanese round-up on Pentium-M machines, linked from the X1000 forums.

Original x1000 forums thread

This test used TMPGenc Plus 2.51 to encode a DV clip to MPEG-2. Their general results showed that the fastest P-M 1.5GHz and 1.6GHz chips took around 2 times real-time to encode the clip, while the low-voltage P-M 900MHz and 1GHz took closer to 3 times real-time. You should take a look at the link because it's interesting to see certain lower-clocked P-Ms machine perform better than higher ones in this test, but here's a sampling of the results:

Thinkpad T40p 1.6GHz - 15.97 fps (real-time playback speed is 29.97fps)
HP NX 7000 1.6GHz - 15.56
Thinkpad X31 1.3GHz - 14.3
Sony Vaio Z1 1.5GHz - 14.3
Thinkpad T40 1.5GHz - 13.66
Fujitsu LOOX T 900MHz - 10 (P5000 series)
Sony Vaio TR 900MHz - 9.17

It would be interesting to see how various P4s and Athlons perform relative to these P-M results. Unfortunately, I'm not sure exactly what TMPGenc settings were used. Also, TMPGenc is known as one of the slower performing encoders in general, so testing it against a different MPEG-2 encoder wouldn't be useful for a comparison.

All P4 reviews with encoder tests have been similar to the one in Attack of the Mega-Notebooks, with nothing as useful as the straight-forward encoding fps numbers from the Japanese P-M roundup.

Again, I suspect the P4 would do substantially better in MPEG-2 encoding. Possibly 20-25%+ better in a 1.7GHz P-M vs. 3GHz P4 test.

Also, I'm not sure that a 1.6 or 1.7GHz P-M will play back the highest resolution, High Definition Window Media video clips smoothly: High Definition Content Showcase. (Especially the 1080p clips like "Coral Reef Adventure") HD video will be more and more common soon (there has already been a Windows Media High Definition version of Terminator 2 released) and it would be nice to know if a P-M machine you buy today will be able to play back future video releases. Remember when you used to need a hardware card to play DVDs on your PC?

Is your sager fast enough to watch this?

Originally posted by: amphibious
If I'm ALWAYS going to have my laptop plugged in and the price and specs of the laptops are similar am I better off going with Pentium 4?

Sorry for going off on my own tangent. To turn this question on its head, does anyone think that a 3GHz P4 would possibly perform worse than a 1.7GHz in anything but battery life? I would say no. If you're looking for a notebook that will be plugged into the wall all the time and won't be moved around much, a P4 will be an equal or greater performer to a P-M and less expensive.

 

[GnomeCop]

I have a Pentium M 1.7ghz centrino laptop

but I don't have a 3 Ghz P4 desktop. My desktop system is a P4 2.53ghz
I can say for sure in day to day tasks the P-M feels faster. Kinda of makes me sad.

 

[bunnyfubbles]

Originally posted by: dnuggett

Sure a 1.7 PM might beable to give a 3.0 P4 a run for its money in a few apps but in others the P4 3.0 will utterly crush the PM.

Ok, so what apps will it get "crushed"? I'm not calling you out just yet but I'd like to see if this is an opinion w/o fact.

guys on my floor have Centrino powered lappys... comparing them side by side when plugged in I'd take my P4 2.8C over their PM machines. But if I actually needed the laptop for class and was lugging it around and running of the batts and nowhere to plugin I'd definately take their PMs. They can definately keep up in pace and crush mine in battery life/portability, but my 2.8 @ 800MHz fsb with 400MHz DDR in dual channel definately dosen't slack. Sure, it's my word, take it or leave it...common sense would tell you I'm right considering that the P4 is still Intel's flagship for the desktop market. I mean if the Centrino were so superior as it is being pimped in here wouldn't we at least see it on a small scale in the desktop world? Oh wait, we do, and its for enthusiasts who want a passively cooled and near silent as you can get system, but you don't hear them talking about extreme performance, just extreme silence.

 

[dnuggett]

Let me see if I can borrow my brother's laptop to serve as sparring (Uniwill N251C2, barton 2500+, kt333, mobility radeon 9000) I think where the P4 beats the Athlon XP it will surely beat the P-M, and viceversa.

I wouldn't make that assumption.

Could you prove mw wrong?

Not really out to prove you wrong at all. The M chip is a great chip, but by my calculations a 1.7 will run to 2.7 in P4 land. So no, a M 1.7 will not hang with a 3.0 P4. But I think anytime someone says a chip will get "crushed" when it is really not that clear I ask for some facts, otherwise we are talking out the wrong end and that can confuse and mislead people.

Bunnyfubles.... you made absolutely no mention of the clock rating of the M chips you are talking about. Without that info, again your statement doesn't mean much.