[cash]

hi everyone!

im planning to buy a powerbook soon (15", 1GHz model) and was wondering if the pb's architecture relies heavily on the system bus? im wondering because the intel centrino system bus is 400Mhz.

i know its very hard to compare the pb's architecture to the centrino one, but is 167Mhz going to be noticeably slower?

thanks!

--
cash

[Lancer409]

i dunno how to compare a centrino to a pb g4
but ... the powermac g5 is MAGNITUDES higher than the powerbooks ... and some people say the bus is a limiting factor on powerbooks .. i dunno

[iMacfan]

I'm not trying to be a troll here or anything, but isn't this question just academic?

The thing is, if you want a Mac, you'll get a mac, even if the processor inside is a Dorito (OK, sorry Wierd Al...). And if you don't need portability, you'll get a desktop, so you are pretty locked in to whatever Apple gives you. As it is, all the PBs have the same bus and nearly the same processor, so I wouldn't worry about it.

David

[cash]

any other comments?

does anyone know if apple/motorola have any plans to make it faster? that is until the g5 pb is released

will the much-rumored dual g4 powerbook be stuck with this 167MHz system bus?

--
cash

[cash]

hi everyone,

but isn't this question just academic?

yeah, i guess it is. i know... if a computer does what you want in a reasonable time, then it'll do

but i am still interested, and if anyone knows anything then that'll make my day

--
cash

[Filburt]

Originally posted by cash:
any other comments?

does anyone know if apple/motorola have any plans to make it faster? that is until the g5 pb is released

will the much-rumored dual g4 powerbook be stuck with this 167MHz system bus?

The existing PowerBook G4s are significantly slower than Pentium-M notebooks. Although slower front-side bus speed is one of the factors, it is one of many reasons. Most significantly, Pentium-M CPUs have much larger L2 cache as well as other optimization. And Pentium-M CPUs use much less power (about half as much as G4).

Although I think PowerBook G4s are slow in comparison, and maybe too slow for demanding tasks (e.g., video editing, large-scale software development, advanced games), it is quite fast for practical situations (e.g., Office, iLife).

P.S. It's been widely speculated that PowerBook G4 will be replaced by end of 2004, with a minor speed bump and possibly a slight boost in front-side bus (to 200 MHz).

[Lancer409]

Originally posted by iMacfan:
I'm not trying to be a troll here or anything, but isn't this question just academic?

The thing is, if you want a Mac, you'll get a mac, even if the processor inside is a Dorito (OK, sorry Wierd Al...). And if you don't need portability, you'll get a desktop, so you are pretty locked in to whatever Apple gives you. As it is, all the PBs have the same bus and nearly the same processor, so I wouldn't worry about it.

David

good point .. i love my rev c. 12incher

[sniffer]

Graymalkin had a input on the topic earlier wich explains it quite good:
http://forums.macnn.com/showthread.p...7A#post2128130

[Dr.Michael]

Originally posted by cash:
hi everyone!

im planning to buy a powerbook soon (15", 1GHz model) and was wondering if the pb's architecture relies heavily on the system bus? im wondering because the intel centrino system bus is 400Mhz.

i know its very hard to compare the pb's architecture to the centrino one, but is 167Mhz going to be noticeably slower?

thanks!

--
cash

I have a Thinkpad with 1.6 GHz centrino and 400 MHz frontside bus.
In everyday life it feels a little bit faster than my powerbook 12 inch/1GHz, 133 MHz bus.

But not so much that its worth to limit yourself to one of the operating systems that are available for that machine.

[phantomac]

When I run Xbench on my PB the memory speed isn't too bad compared to the PM G5. Honestly.

The only bottleneck on my machine is the 4200 RPM harddrive. That's going to get swapped for a 7200 RPM Hitachi once the larger sizes become available.

[danny_gasperini]

The debate about how bus speed directly impacts on performance on the current Mac hardware has been raging for ages. Traditionally, people thought that upon releasing the PowerMac G5 hardware, with bus speeds increasing from 167MHZ (in the dual 1.42GHz PowerMac) to 800MHz, dual 900MHz and Dual 1GHz would make a MASSIVE difference. The reality has been, however, that the increase in overall system speed has been no where near an order of maginitude proprtional to the bus speed increase. Take a look at the following tests, for example:

http://www.barefeats.com/imacg5.html

This compares a variety of the current Mac hardware, with a variety of applications, so as not to bias one system against another.

If we take a look at a G4 system with a 167MHz bus and compare it against a G5 system with a 533MHz bus and use similar CPU clock speeds (ie, a PB 17" G4 @ 1.5GHz/Radeon 9700 and an iMac 17" G5 @ 1.6GHz) what do we see?

1. Motion Render RAM Preview - PowerBook is almost twice as fast. Why? The superior graphics card is almost definately the reason here.

2. Motion Playback Without Render - PowerBook is about 40% faster. Why? The superior graphics card is almost definately the reason here.

3. iMovie Render Soft Focus Effect - iMac is about 6% faster. Why? This small speed difference could easily be accounted for by the 100MHz higher clock speed on the G5 processor alone.

4. Cinebench 2003 CPU Render - iMac is about 39% faster. Why? Cinebench makes heavy use of the memory and CPU bandwidth, which are higher on the iMac.

5. FileMaker scripts - iMac is about 4% faster. Why? Once again, probably the 1600MHz v's the 1500MHz difference.

6. Halo 1024x768 Time Demo - PowerBook is about 15% faster. Why? Halo is a GPU intensive game. The Radeon 9700 outshines the GeForce FX5200. It's even more impressive than this figuire suggests given that the image generated on the ATI card is superior to the nVidia card in Halo.

7. UT2004 Botmatch test - PowerBook is faster by about 5%. Why? Not sure. Theoretically the iMac should be slightly faster on this since it has a slightly higher clock speed. The UT2004 Botmatch tests are VERY CPU dependant.


Conclusions: The 3 x higher bus speed on the G5 does not mean it is 3 x faster than the G4 at the same clock speed in real world tasks. In fact it doesn't seem to make ANY SUBSTANTIAL difference on most tasks. The PowerBook's much faster graphics give it a significant edge, however, in current high end applications like Motion, and games like Halo. The bus speed also seems to make much less difference than the processor clock speed.

[Fabiano007]

danny_gasperini

Thanks!!!

Great Post!

[Filburt]

Originally posted by danny_gasperini:
Conclusions: The 3 x higher bus speed on the G5 does not mean it is 3 x faster than the G4 at the same clock speed in real world tasks. In fact it doesn't seem to make ANY SUBSTANTIAL difference on most tasks. The PowerBook's much faster graphics give it a significant edge, however, in current high end applications like Motion, and games like Halo. The bus speed also seems to make much less difference than the processor clock speed.

Very true and good observation. However, iMac G5s are significantly cheaper than PowerBook G4s and aside from GPU, they perform faster. PowerBook G4s rank poor in cost-to-performance ratio.

[danny_gasperini]

Originally posted by Filburt:
Very true and good observation. However, iMac G5s are significantly cheaper than PowerBook G4s and aside from GPU, they perform faster. PowerBook G4s rank poor in cost-to-performance ratio.

Yes, the iMacs are cheaper than the PowerBooks. However, the technology involved in minimising the components for portable use will almost always be more expensive than desktop technology; where constraints on size, heat, and power comsumption are less. What price would someone be prepared to put on portability? That is the question.

The PowerBooks are also aimed at a different market segment than the iMacs - Pro v's Consumer. I was using the example of the iMac, soley to illustrate the G5 processor and system bus compared to the PowerBooks processor and system bus. The results indicate no statistically significant difference in real world tasks when MHz are corrected for each processor.

[danny_gasperini]

UPDATE>

www.barefeats.com has actually just updated their graphs due to people doubting their results. After testing and retesting, even using different testing procedures and benchmark utilities, the PowerBook has actually increased it's advantage over the iMac G5/1.6 and even the iMac G5 1.8, in the UT2K4 Botmatch tests.

[Pierre B.]

Originally posted by danny_gasperini:
Traditionally, people thought that upon releasing the PowerMac G5 hardware, with bus speeds increasing from 167MHZ (in the dual 1.42GHz PowerMac) to 800MHz, dual 900MHz and Dual 1GHz would make a MASSIVE difference. The reality has been, however, that the increase in overall system speed has been no where near an order of maginitude proprtional to the bus speed increase.

Yes, it is nearly an order of magnitude faster in certain situations. In common tasks where the system is not pushed hard, the only difference is indeed what the clock speed will let you assume. But when MASSIVE amounts of data need to travel between the processor and the memory, the G5 FSB can make a MASSIVE difference.

Here is an example. Look at the beginning of the discussion where a test is performed on a dual G4 1 GHz and a dual G5 2.5 GHz (rendering of a 600 MB video file). You will see that the G5 completes the test in 3-4 minutes, while the G4 needs a little less than an hour! So, in this test the G5 is not 2.5 times faster, as the clock speed indicates, but almost 15 times faster.

In tests where the system bus of the G4 is not filled up completely (and consequently, the bus of the G5 is way under-utilized), the performance differences are mostly due to L2 cache, pipeline length (here the G4 has an advantage), clock speed and graphics subsystem.

The G5 shines only in very heavy calculations (especially those involving the FPU), where a G4 would choke to death, by the need either of addressing huge amounts of memory or moving huge amounts of data. Otherwise, the differences are small, not to say that the G4 can be in advantage in some cases.

[sniffer]

Great inputs. I'll guess this sums up the G5 advance is in general within video editing and music making/sampling, both which costs a lot of bandwidth. There is also the clock speed difference, but there is still the barrier to get the G5 cool enough to fit into a PB encloser, so comparing the PB 1.5 with iMac 1.6 sounds fair enough to me.

[68k33]

Interesting, that in today's world the focus is mostly on hardware specifics.

But: What about the software-coding ? High-End Assembler-Optimization for the specific architecture ?

I will just remind you of an example: Preview in OS X 10.2 and 10.3. I use a Powerbook 15" 800 MHz / DVI (1 MB Level 3 Cache), now over two years "old". But even on this machine Preview flies in 10.3 that it is amazing. It Even outperforms most PDF-Readers out there.......

[icruise]

Originally posted by Filburt:
Although I think PowerBook G4s are slow in comparison, and maybe too slow for demanding tasks (e.g., video editing, large-scale software development, advanced games), it is quite fast for practical situations (e.g., Office, iLife).

I've never used a Pentium M notebook, so I can't comment on the relative speed difference, but I think it is totally wrong to say that current Powerbooks aren't up to tasks like video editing or software development. People (myself included) have been editing video with powerbooks for years. Does it make sense that the fastest powerbooks ever would suddenly be too slow for the task?