Powerbook 15" @ 1.5GHz

[alisajid]

I had to let go of my 1GHz 15" Powerbook because it felt too slow to me (e.g. @ running eclipse). Since the 1.5GHz 15" Powerbook came out, I've become more hopeful, but I'd like to know what kind of performance I can get out of it compared to lets a 1.6 GHz Pentium M in an IBM T41 for tasks such as compiling Java and C source and web browsing.

 

[hopejr]

I'm sure it's faster than the similar clocked P-M.
I noticed that eclipse was a bit slow on my 1GHz iBook. I thought I was just running too many programs (i did have 6 other programs running, not to mention all the processes running in the background).

 

[alisajid]

The reason why I'm a little hesitant is that my bro's eMac 1.25GHz G4 seems kind of slow for day to day tasks. I would love to switch back, but I need at least near equivalent performance as my current lappy (t42 @ 1.7 dothan).

 

[edtsui]

I love my 1.33ghz 15in but I have to admit, its not gonna break any speed barriers. Its good for the basic tasks but if you're hardcore you might want to stick with a P-M.

 

[randumb]

Originally posted by: hopejr
I'm sure it's faster than the similar clocked P-M.
I noticed that eclipse was a bit slow on my 1GHz iBook. I thought I was just running too many programs (i did have 6 other programs running, not to mention all the processes running in the background).

I'm pretty sure the Pentium M beats the G4 clock for clock.

 

[imported_Lucifer]

Originally posted by: randumb

Originally posted by: hopejr
I'm sure it's faster than the similar clocked P-M.
I noticed that eclipse was a bit slow on my 1GHz iBook. I thought I was just running too many programs (i did have 6 other programs running, not to mention all the processes running in the background).

I'm pretty sure the Pentium M beats the G4 clock for clock.

And how would this be possible? On AltiVec enabled tasks, the G4 processor can produce data at 128 bits. The P-M produces data at 32 bits. Want to know what AltiVec is? Do a search on it. AltiVec is what made my 1GHz eMac whoop my friends AMD XP 2000+ running at 2.07GHz.

Meaning, the 1.5GHz Powerbook will outperform your 1.7GHz dothan in tasks that are altivec enabled, such as many filters in photoshop, image sizing in photoshop and QuickTime compression.

My 1GHz eMac is fast as hell to me. I love it. Its got plenty of speed to do what I mostly do.

 

[RichieZ]

its slow, i have the 1.2g ibook and its significatly slower than my 1.4 Pentium M thinkpad X31. I love OS X

 

[littlezipp]

Using a 1.33ghz 12" powerbook, it is slower than a Dell 1.3ghz PM.
Just my 2 cents

 

[hopejr]

I rather call AltiVec the Velocity Engine, but that's just me Quite a few programs don't take advantage of the speed from the Velocity Engine, and that can turn some people off if they don't get the chance to see it work.
My 1GHz iBook G4 is only slightly slower than my old 2.2GHz P4-M. In fact, if this had as much memory as the P4-M, is might actually be the same (I only have the stock 256MB where the P4-M had 512MB).

 

[alisajid]

I don't care as much about objective performance metrics like clock-for-clock comparisons. What matters to me is performance in the apps I use across platforms, e.g. Eclipse and OpenOffice. More importantly I don't do any graphics related or any floating point intensive applications. Even so, with respect to Altivec and 128 bit instructions, SSE provides 128 bits registers on the intel platform anyway, either way it doesn't matter, if it 'feels' faster is all that really matters.

 

[imported_Lucifer]

SSE does not provide 128 bit registers. SSE produces data at 32 bits. SSE2 produces data at 128 bits. The P-M is not SSE2.

My 1GHz eMac feels a whole lot faster than my uncles 1.2GHz AMD Athlon T-bird. For what you are doing, the Powerbook will feel fast at it.

 

[RichieZ]

Originally posted by: alisajid
I don't care as much about objective performance metrics like clock-for-clock comparisons. What matters to me is performance in the apps I use across platforms, e.g. Eclipse and OpenOffice. More importantly I don't do any graphics related or any floating point intensive applications. Even so, with respect to Altivec and 128 bit instructions, SSE provides 128 bits registers on the intel platform anyway, either way it doesn't matter, if it 'feels' faster is all that really matters.

its feels much slower, i run eclipse on both my 1.2g ibook and also on my 1.4 pentium M. Safari also feels slower than IE, scrolling webpages is pretty slow

i only got a mac b/c i wanted OS/X. If you want performance you need to get a G5 desktop or get a Pentium M

 

[imported_Lucifer]

Well, not sure what is wrong with your iBook, because my 1GHz eMac runs very fast.

 

[Burbot]

Thin Lizzy

And how would this be possible? On AltiVec enabled tasks, the G4 processor can produce data at 128 bits. The P-M produces data at 32 bits.

SSE registers are 128 bit, so G4 has no advantage here.

Meaning, the 1.5GHz Powerbook will outperform your 1.7GHz dothan in tasks that are altivec enabled, such as many filters in photoshop, image sizing in photoshop and QuickTime compression.

If a task can be "altivec enabled", in majority of cases it can be "SSE enabled".

SSE does not provide 128 bit registers. SSE produces data at 32 bits. SSE2 produces data at 128 bits. The P-M is not SSE2.

First, SSE *does* provide 128-bit registers. Second, Pentium M *does* support SSE2. Link

Why don't you quit spreading lies and disinformation and go back to worshipping brushed aluminum?

 

[imported_Lucifer]

Did you not read what I said? The G4's registers are 128 bit, so why do you say the G4 has no advantage?

Well, if a task is SSE enabled, then why didnt it help while using Photoshop. My friends AMD 2000+ has SSE, and he still lost to my 1GHz Mac. So doesnt look like SSE was enabled.

Well, we will have to see what is faster. The 1.5GHz G4 or the 1.7 P-M

Why don't you quit spreading lies and disinformation and go back to worshipping brushed aluminum?

Got a problem with brushed aluminum? Go back to worshiping plastic then.

 

[Commodus]

I would wait until Apple (hopefully) updates the PowerBooks in the early part of 2005. You might not just get a clock speed boost - if they use the upcoming G4s, you'd get a faster FSB (200 MHz), more L2 cache (1 MB), and a clock speed boost.

 

[alisajid]

Thanks to all who shared their Eclipse on (i|(Power))book experiences. I guess I'll stick with Gentoo on the T42 till the new Powerbook's show up.

 

[imported_Lucifer]

Nice choice alisajid. Do not believe Burbot about SSE being 128 bit registers. My brother, who does his work on PC, told me that SSE is 32 bit registers. SSE2 is 128 bit registers.

Good luck alisajid with your choice. Hope everything works out fine for you.

 

[ProCoolingKevin]

Honestly, I have an iBook 1.2 ghz G4 w/ 768mb RAM. It's pretty zippy in OSX. But don't buy a Mac if you're going for pure speed. You'll be dissapointed. I got swayed cuz I love OS X, the iLife suite of apps, and X11 integration to run linux apps.
-Kevin

 

[Burbot]

Intel® Architecture Optimization Reference Manual
Page 36.


Streaming SIMD Extensions of the Pentium III Processor

The Streaming SIMD Extensions of the Pentium III processor accelerate
performance of applications over the Pentium II processors, for example,
3D graphics. The programming model is similar to the MMX? technology
model except that instructions now operate on new packed floating-point
data types, which contain four single-precision floating-point numbers.
The Streaming SIMD Extensions of the Pentium III processor introduce new
general purpose floating-point instructions, which operate on a new set of
eight 128-bit Streaming SIMD Extensions registers.

(empasis mine)
There is also a nice figure on page 38 for those who still believe that this is a typo.



So, pretty much every modern processor can produce information in "128 bit chunks" (starting from Athlon XP and Pentium III). G4 is nothing special.

As for Photoshop on machine X being faster then on machine Y, there are a *lot* of factors (memory size, memory clock rate, hard disk speed, filesystem fragmentation, software churning in background) that affect performance. If you would like to compare performance, take two systems with identical RAM and disk configuration, and make sure Photoshop has separate drive(s) just for scratch files. Ad-hoc comparisons have too many areas where things might get screwed up.

 

[imported_Lucifer]

Well, I guess you are right. But that doesnt make a computer with SSE faster than the G4. If it were, my 1GHz eMac would have lost to my friends AMD 2000+, which uses SSE.


Here is what is special about AltiVec compared to SSE, MMX, and SSE2

Altivec:

32 separate Registers

128 bits per register

No interference with FP registers

no context or mode switching

max throughput: 8 Flops / cycle

SSE, MMX

8 MMX registers shared with the FPU, 8 for SSE

64 bits per mmx register, 128 bits per xmm register

MMX stalls the FP registers

context switching required for MMX

max throughput: 2 Flops / cycle

Also, I have a friend who is both Apple and Microsoft, but is hardcore PC right now. He will even tell me the G4 is better than SSE, as he did before, while I kicked his AMD's ass!!

Interesting article on AltiVec.
Interesting article on SSE, MMX

Compare them and tell me what you think is better.

EDIT: You know, read it all. Read all the way to the conclusion.

 

[Burbot]

Your article states that AltiVec is better then SSE1. All modern processors from Intel and AMD support SSE2, which was a big improvement to SSE1. Why don't you find an article that compares current tech to current tech, that is AltiVec to SSE2?

As for Apple "max throughput" figures, they deserve as much scrutinity as any Apple benchmark. Those are only valid for fused multiply-add, and are fairly decieving, since under majority of workloads G4 memory bus will be unable to keep up with CPU. And since the thread creator cares a lot more about integer and logic performance anyway, all that SIMD talk is pointless anyway.

 

[vexingv]

i have a 1.33 g4 12" PB, and ms word on it seems slower than it does on my 1700+ athlon. it just may be MS' coding or implementation.
and as someone else mentioned before, you should def. wait until at least mid january or so when the next Mac expo/event will take place in case a new product is launched.

 

[imported_Lucifer]

Read This

Quoted from that page.

As for SSE2 vs. Altivec, SSE2 is a toy by comparison. Its architecture does not offer the range of generalized high precision capability that the altivec instruction set does. It is filled with bandwidth limitations, particularly its tiny number of harder to use registers that make it nearly impossible to keep the pipeline full, and it is capable of basically no parallelism whatsoever with the regular FP unit on the processor (which means it must start and stop each unit to switch back and forth, and the lack of generalization makes this an excruciating performance penalty). The small number of registers in particular makes the P3 a better scientific computing processor than the P4 for real world applications because the P4's pipe is too deep to keep it filled. This can be graphically demonstrated with fully optimized applications that force significant branching on real world data.


The newer PPC7450 series machines have even better vector performance than their predecessors, for a ruthless extremely optimized cross-platform demonstration, try running distributed.net's RC5 cracker on a G4 against anything else. The dual 1ghz G4 cracks more than 25 million keys a second. That is more than an order of magnitude faster than a 1.5ghz P4 from Dell does on the exact same job. As for P3, it does better than the P4, but still runs three times slower per clock than the 7400 G4. (The newer 7450 is substantially 3x faster than clock). A 1 Ghz P3 outperforms the 1.5 Ghz P4 by a sound margin. This application is a good demonstration of the problems with the P4's ridiculously deep pipes and crappy registers.

As for MS Office for Mac, yes, it is slower than the PC version.

 

[RichieZ]

stop being a mac fanboy