MacSpeedZone just released the speed tests for the 900MHz. Since there's been a lot of talk about iBooks vs FireBooks, I thought I'd post the link:
900 MHz iBook Tests
Parsa
900 MHz iBook Tests
Parsa
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
iBook 900MHz speed tests
- Thread starter Parsa
- Start date
- Sort by reaction score
i'm confused by the results. why aren't the values expressed in time... how can a percentage give people an accurate understanding of the performance.
for example, in the Finder File Search test, they show the result for the iBook 900 as 70%, then they say underneath that, that "Only 40% of the processing power of the 900 MHz iBook was used ... leaving 60% looking for something to do".
what exactly does the 70% represent, if only 40% of the processing power was the amount used? i would be interested in understanding the way they tested these so i can better understand how the iBook 900 stacks up.
if anyone can explain their testing methods, that would be very helpful.
for example, in the Finder File Search test, they show the result for the iBook 900 as 70%, then they say underneath that, that "Only 40% of the processing power of the 900 MHz iBook was used ... leaving 60% looking for something to do".
what exactly does the 70% represent, if only 40% of the processing power was the amount used? i would be interested in understanding the way they tested these so i can better understand how the iBook 900 stacks up.
if anyone can explain their testing methods, that would be very helpful.
Actually if you go through test by test, the iBook 900 does pretty darn good against the Powerbook 12" 867. The only place I can see it gets killed is Alti-Vec specific tasks. Even MP3 encoding, and DV importing are good. Some places the iBook even took out the 12" Powerbook and killed the older Powerbook 500's and 550's. I guess sometimes Mhz does matter. Just thought this test is interesting since I have heard it suggested that the iBook should be a G4 with a lower clock speed.
what is real
I have both in front of me right now and i can say that from just general use that the 867 G4 is like a desktop machine with the bits shaved off! the iBook just gives me more spinning beachballs from day to day (only had the 900 a 5 days though - and is a german version - not that this affects it!!). I really like the 867 and Im a graphics intensive user - maybe the 900 is ok for general office or programming use but not for me!! I hate to say this but games really dont do as well on the 900 too - but we do use Maya/photoshop/illustrator and thats seems to have a big difference on a g4 chip!!
I have both in front of me right now and i can say that from just general use that the 867 G4 is like a desktop machine with the bits shaved off! the iBook just gives me more spinning beachballs from day to day (only had the 900 a 5 days though - and is a german version - not that this affects it!!). I really like the 867 and Im a graphics intensive user - maybe the 900 is ok for general office or programming use but not for me!! I hate to say this but games really dont do as well on the 900 too - but we do use Maya/photoshop/illustrator and thats seems to have a big difference on a g4 chip!!
RTFA
iBook 900 14" with 640MB of RAM
100% = performance of a dual proc 1GHz PowerMac
If dual G4 1GHz takes 1 minute, and the iBook is 50% as fast; take 1minute and divide by 0.5 (translates into 50% as fast as dual proc 1GHz G4).
For the money, if you only lightly use Altivec, you are better off to buy the iBook 900 than to buy the 12" PB 867 G4. This is true even for Excel/Word macros, non-Altivec Photoshop, CD burning (newer optical drive), Stuffit decoding, etc etc.
For most use, incuding some casual Photoshop work, and you will be getting same / better performance, for a much better price!
iBook 900 14" with 640MB of RAM
100% = performance of a dual proc 1GHz PowerMac
If dual G4 1GHz takes 1 minute, and the iBook is 50% as fast; take 1minute and divide by 0.5 (translates into 50% as fast as dual proc 1GHz G4).
For the money, if you only lightly use Altivec, you are better off to buy the iBook 900 than to buy the 12" PB 867 G4. This is true even for Excel/Word macros, non-Altivec Photoshop, CD burning (newer optical drive), Stuffit decoding, etc etc.
For most use, incuding some casual Photoshop work, and you will be getting same / better performance, for a much better price!
Re: what is real
the games thing just makes no sense. Mhz matters for games, not G4 vs. G3. And they both have 32 megs of VRAM on roughly comparably video cards.
as for the speed difference you find, the applications you use explain it because you use altivec aware apps. Those apps are the only ones where the PB crushed the ibook.
the games thing just makes no sense. Mhz matters for games, not G4 vs. G3. And they both have 32 megs of VRAM on roughly comparably video cards.
as for the speed difference you find, the applications you use explain it because you use altivec aware apps. Those apps are the only ones where the PB crushed the ibook.
let th figures speak for themselves
I agree on paper the 900 looks good - thats why we got one - but in reality i just cant get the same out of them both - does anyone else have these 2 machines both with 512Ram in them? maybe theres an X factor - our 867 was second hand and cost the same as the 900iBook new - but personally i would prefer 2 x 867!!
I agree on paper the 900 looks good - thats why we got one - but in reality i just cant get the same out of them both - does anyone else have these 2 machines both with 512Ram in them? maybe theres an X factor - our 867 was second hand and cost the same as the 900iBook new - but personally i would prefer 2 x 867!!
Originally posted by cb911
for example, in the Finder File Search test, they show the result for the iBook 900 as 70%, then they say underneath that, that "Only 40% of the processing power of the 900 MHz iBook was used ... leaving 60% looking for something to do".
what exactly does the 70% represent, if only 40% of the processing power was the amount used? i would be interested in understanding the way they tested these so i can better understand how the iBook 900 stacks up.
if anyone can explain their testing methods, that would be very helpful.
As previously explained the iBook did the test 70% as fast as their "standard" dual cpu system. In other words, it took 1/.70 times as long.
During 60% of the time it took, the processor was running in place, waiting for something like data from the hard drive, memory, etc.
Which is good that finder doesn't hi-jack all the system resources. Unlike Windows 98, when I do a search (Outlook usually) at work, everything slows to a crawl until it is done
Yeah, even Windows 2000 has that...
It's weird, since Windows 2000 should really be a good multitasker and multithreader, but on simple things like that (e.g. performing a search in the Explorer) it sometimes crawls to a near halt. Very annoying; never had that in OS X. I'm also quite curious how my somewhat older iMac with an 800MHz G4 stacks up against this new 900MHz iBook. I'd also like to know where the graphics cards stand; is the GeForce 2MX better or lesser than the Radeon 7500 (both have 32MB)?
Originally posted by yzedf
Which is good that finder doesn't hi-jack all the system resources. Unlike Windows 98, when I do a search (Outlook usually) at work, everything slows to a crawl until it is done
It's weird, since Windows 2000 should really be a good multitasker and multithreader, but on simple things like that (e.g. performing a search in the Explorer) it sometimes crawls to a near halt. Very annoying; never had that in OS X. I'm also quite curious how my somewhat older iMac with an 800MHz G4 stacks up against this new 900MHz iBook. I'd also like to know where the graphics cards stand; is the GeForce 2MX better or lesser than the Radeon 7500 (both have 32MB)?
Re: Re: let th figures speak for themselves
You can't. 384 or 640 only. If Apple pulled their head out... you could.
Bengt77:
The Radeon is a better card but we don't know what speed Apple has it clocked at in their laptop. At standard desktop speeds and when using PC drivers, a Radeon 7500 is much faster than any GF2MX. Honestly I don't pay enough attention to Mac benchmarks to tell you if anyone has numbers on those cards for Macs.
the reason gaming performance is better on the PB 12" is because of the system/memory bus. The only reason the iBook is faster in some tests is because it has twice the L2 cache then the G4. (512K vs 256K) I'd rather have 512K on chip cache vs 256K + 1MB L3 (If only the G4 could have 512K). Remember that on chip L2 cache runs at the speed of the processor (in the case of the ibook, 900mhz) Since it has twice the cache, it can actually perform certain things that will fit in the cache faster because it doesn't have to access the slower system bus. When you get out of the cache, the PB will be faster in most cases because of its faster system/memory busses. Also have to take into consideration the G3 still has the shorter, more efficient 4 stage pipeline vs the 7 stage pipeline of the newer G4's.
The people who thought that the last revs of PowerMacs were getting the 7457s were delusional. According to Motorola's press releases, the 7457s are available for sampling now and will be ready to ship in Q3 2003. So if the 970s aren't ready and the 7457s are, look for the August 2003 PowerMacs to sport another G4.Originally posted by Funkatation
I realize that... but who knows if we'll ever see those
Even if the 970 is ready for August, the 7457 may find it's way into the PowerBooks.
Actually, L2 cache can run at 1/2 speed or full speed of the processor.
Either way, the 12" PB got neutered when they didn't include the 1MB of L3 cache.
it only runs at half speed when its off the die i.e backside cache. when its on die cache, it runs at full processor speed. (all current processors use the latter type of cache)
Well, I didn't find this article that informative, as it really left a lot of details out. For example, it identifies the PowerBooks simply by clock speed, not by model. We know that there are performance differences between the 12" 867MHz G4 and the 15" 867MHz G4 models, just as there are performance differences between the 15" 1GHz G4 and the 17" 1GHz G4 models. So, we know far too little about how these tests were run to really know anything for sure...
bb0ys:
Funkatation:
Cause Bengt77 was wondering about how the iBook compared to his iMac.
Funkatation:
The speed of backside cache is configurable, and includes speeds such as 1/2, 1/3, 1/4 and 1/6, amoung many others. Apple has used a variety of fractional speeds for their L3's, with the original 733-G4 probably being the last with true 1/2-speed SDR backside cache. My dual 800-G4 has 1/4 speed SDR L3's, and the original dual 1gig had 1/4 speed DDR L3's (which makes them sort of 1/2 speed). I don't think that Apple has ever offered a backside cache with an effective clock much higher than 500mhz. The dual 1.25's probably have 1/5 speed DDR L3's and I bet the 1.42 has something like 1/6 speed (but DDR of course).
They're talking about the GF2MX because bengt77 asked about the current ibook as compared to his 800 mhz imac which has a GF2MX.
I was saying that on die cache was always full speed, and yes backside cache (L2 or L3) is configurable depending on the speed of the SRAM.