Interesting points of this (from a layman's perspective):The Jet3D test only needed 1MB of memory to run, so system bandwidth may not have come into play(?).
The G4, when averaged to MFLOP/MHz, is on par with the G5, pretty much.
The author comments that if the test were compiled for the G5, a 20% boost in performance is feasible.
His conclusion is basically that a G5 running unoptimized code is about 20% faster than a G4 on a per MHz basis. That is pretty darn good! Remember that the P4 was slower on a per MHz basis than the PIII. G5's are faster per clock cycle AND have more clock cycles.
Also very important is the fact that his test does not measure bus speed. This is good for getting an idea of the CPU speed, but a real world test would also take bus speed into account and this is where the G5 rules.
This is a very important test for all those people who worry that a G5 won't be able to run a program written for a G4 very quickly. Your fears are unfounded and you don't have to worry. The G5 rules.
Concerning Photoshop -- The best and most successful products are those that truly fill a need. The majority of pro designers in the world are not techies. They are very good working with Photoshop and the page layout program of their choice.
They have always been drawn to the Mac based on the OS elegance (that they like to feel associated with) and ease of use. Most are lost once they leave the comfort of their day to day graphics software.
Many still have put up with the relatively slow speeds of recent pro towers, but many more, by choice, or by IT directive, have migrated to the "more bang for the buck" PC's.
But their "need" is still there. Photoshop on a G5 will again become the solution of choice because the OS is even more elegant (from a users point of view), and the G5 Mac finally has bragging rights.
Adobe won't want to disappoint once the G5 hits the main stream.
Hrmmph. Still too early for solid G5 benchmarks. I hate seeing benchmarks that run unoptimized code on a G5, and then show a Pentium 4 running optimized code that hands the G5's ass to it. Not cool.
It's not enough for Apple to simply "catch up", or "close the gap", and it never has been; Apple has always offered a premium product with state of the art processing power, save for the Moto debacle, and this performance attribute constitutes a large part of Apple's identity. Thus, keeping with tradition, the Powermac G5 must be unequivocally faster than any Pentium 4 based Wintels. Only a new performance gap that puts Wintels on the butt end of the stick will trigger the sort of mass migration to OS X that Apple needs for vitality. As Steve Jobs would say, we need Powermacs that are so fast they "make us shit in our pants".
Here's to hoping that the G5 dominates Wintels right out of the gate, so when Apple unleashes the Dual 3 GHz Powermac G5 upon the masses it will so totally destroy Wintels that Arstechnica will be forced to pull the plug on their Battlefront forum for wont of anything to discuss.
Hrmmph. Still too early for solid G5 benchmarks. I hate seeing benchmarks that run unoptimized code on a G5, and then show a Pentium 4 running optimized code that hands the G5's ass to it. Not cool.
It's not enough for Apple to simply "catch up", or "close the gap", and it never has been; Apple has always offered a premium product with state of the art processing power, save for the Moto debacle, and this performance attribute constitutes a large part of Apple's identity. Thus, keeping with tradition, the Powermac G5 must be unequivocally faster than any Pentium 4 based Wintels. Only a new performance gap that puts Wintels on the butt end of the stick will trigger the sort of mass migration to OS X that Apple needs for vitality. As Steve Jobs would say, we need Powermacs that are so fast they "make us shit in our pants".
Here's to hoping that the G5 dominates Wintels right out of the gate, so when Apple unleashes the Dual 3 GHz Powermac G5 upon the masses it will so totally destroy Wintels that Arstechnica will be forced to pull the plug on their Battlefront forum for wont of anything to discuss.
Apple, don't fsck up!
G5 + optimised soft for the G5 + panther = kick ass computer.
After Effects is more suited towards motion graphics...
Shake is definitely a compositing tool...
There is some overlap in what they can do, and for the longest time After Effects was the standard in lower-priced desktop guerilla compositing... As with most Adobe products, plug-ins make to difference at the end of the day...
But Shake is much more suited for pure compositing work, stuff where you already have all of the elements needed for the shot, you just need to comp them together...
Plus, it is an Apple product...
Right now Adobe (and the boutique shop 'the Orphanage'...) is on my shitte list after releasing that ad dumping on Apple computers in LOTS of trade magazines... Bastards!
...Right now Adobe (and the boutique shop 'the Orphanage'...) is on my shitte list after releasing that ad dumping on Apple computers in LOTS of trade magazines... Bastards!
Just wern't right, Man!
I just subscribe to DV magazine for the heck of it (I'm a DV hobbyist), and the first time I open my first issue, I see that fricken' ad. Can't wait for a new AE bake-off with a dual G5.
Well if you take the percentage change from the "dual processor AE" article (45.46 to 52.4%) and then turn your attention to Chalie whites article found here:
and adjust the times of the mac by multiplying the Dual 1.4 G4 system times by .56 ( 56% of the time recorded by Mr. White) you get a whole different impression of the Dual G4's speed.
minutes: seconds, winner in boldface type
Apple Power Mac G4 Dual 1.25GHz with 1GB DDR RAM
$3,589
Dell Precision Workstation 350
Intel P4 3.06 GHz,
1GB PC1066 RDRAM
$2964
(Estimated times in parenthesis)
1. After Effects: Simple Animation
:14 (:07.84 sec)
:07
2. After Effects: Video Composite
1:25(~50.4 Sec)
:54
3. After Effects: Data Project
3:47 ( ~2:10)
2:05
4. After Effects: Gambler
:43 (24.08)
:29
5. After Effects: Source Shapes
7:06 ( ~4:00)
4:14
6. After Effects: Virtual Set
8:15 (~4:40)
4:24
That would make the Dual G4 Stand up quite nicely.
The Mac speeds would be the first set of number (as reported by charlie white) with the modified speed in parenthesis. The Dell P4 speeds are the second number
Using the Night Flight benchmark on the unoptimized After Effects 5.5, the G5 was 39% faster. Performance tests included Photoshop and Cubase, and a technology demonstration of the G5 masking video.
Well, I guess since the NDA i signed is not necessary anymore since they introduced the new version of premiere and after effects, I guess I can tell you some things I know. I work at one of the Art Institutes, and we wanted to cut back on our license spending for adobe products (we just buy blanket licenses, and we even have licenses for apps that we have never even installed) and obviously adobe gets a huge check from our school ever month. So Adobe was worried about losing a huge source of revenue, so they schedualed a real promp demonstration to show off new versions of premiere and after effects about 3 months ago obviously before they announced (which required my NDA). I was kinda curious why they chose to do this in a PC lab (I'm a Mac tech), and then they finally explained that Premiere was going PC only. The reason they gave was that the current (the then current G4) had way too slow of a bus speed and they couldn't get any really usable performance out of it. They said the reason why Final Cut Pro was able to get good performance on G4 was that Apple has such deep hooks into the Quicktime layer of OS X, but that they haven't really shared that low level access to Developers yet. But I was also very interested when he said that they had started seeing new test machines from Apple that would change all that and they were very excited about, but when I asked him at the time if he could talk anymore, his answer was "well, unfortunately i have to sign my own NDA's." But he did say that there is a very nice chance to see Premiere on the Mac again in the future, but they feel that they have to re-earn the industries respect with premiere first, since he openly admitted version 6.5 was horrible. (Version 7.0 was completely re-written from the ground up). And oh yeah, their is a programmer who freelances and does all the color correction for all the pro video apps. The same guy who did it for final cut did it for the new version of premiere and a few other applications as well. it's pretty interesting stuff. Well, hope you found this of any interest.
I wouldn't worry too much about "cooling off noises" from Adobe or Quark ? mainly just to increase pressure on Apple for financial/programming assistance in Mac OS X development, to extract concessions etc. So long as there's a business case for Mac versions, they will continue to be developed. With Premiere, Final Cut Pro is so good that the medium term revenue outlook on the Mac for the product is poor, given the performance ratio versus FCP, thus meaning it gets knifed, for now if what The Letter Exchange is saying is true
Comments
Originally posted by scottiB
Interesting points of this (from a layman's perspective):The Jet3D test only needed 1MB of memory to run, so system bandwidth may not have come into play(?).
The G4, when averaged to MFLOP/MHz, is on par with the G5, pretty much.
The author comments that if the test were compiled for the G5, a 20% boost in performance is feasible.
His conclusion is basically that a G5 running unoptimized code is about 20% faster than a G4 on a per MHz basis. That is pretty darn good! Remember that the P4 was slower on a per MHz basis than the PIII. G5's are faster per clock cycle AND have more clock cycles.
Also very important is the fact that his test does not measure bus speed. This is good for getting an idea of the CPU speed, but a real world test would also take bus speed into account and this is where the G5 rules.
This is a very important test for all those people who worry that a G5 won't be able to run a program written for a G4 very quickly. Your fears are unfounded and you don't have to worry. The G5 rules.
Are After Effects and Shake competing products?
Which is better?
They have always been drawn to the Mac based on the OS elegance (that they like to feel associated with) and ease of use. Most are lost once they leave the comfort of their day to day graphics software.
Many still have put up with the relatively slow speeds of recent pro towers, but many more, by choice, or by IT directive, have migrated to the "more bang for the buck" PC's.
But their "need" is still there. Photoshop on a G5 will again become the solution of choice because the OS is even more elegant (from a users point of view), and the G5 Mac finally has bragging rights.
Adobe won't want to disappoint once the G5 hits the main stream.
It's not enough for Apple to simply "catch up", or "close the gap", and it never has been; Apple has always offered a premium product with state of the art processing power, save for the Moto debacle, and this performance attribute constitutes a large part of Apple's identity. Thus, keeping with tradition, the Powermac G5 must be unequivocally faster than any Pentium 4 based Wintels. Only a new performance gap that puts Wintels on the butt end of the stick will trigger the sort of mass migration to OS X that Apple needs for vitality. As Steve Jobs would say, we need Powermacs that are so fast they "make us shit in our pants".
Here's to hoping that the G5 dominates Wintels right out of the gate, so when Apple unleashes the Dual 3 GHz Powermac G5 upon the masses it will so totally destroy Wintels that Arstechnica will be forced to pull the plug on their Battlefront forum for wont of anything to discuss.
Apple, don't fsck up!
Originally posted by Junkyard Dawg
Hrmmph. Still too early for solid G5 benchmarks. I hate seeing benchmarks that run unoptimized code on a G5, and then show a Pentium 4 running optimized code that hands the G5's ass to it. Not cool.
It's not enough for Apple to simply "catch up", or "close the gap", and it never has been; Apple has always offered a premium product with state of the art processing power, save for the Moto debacle, and this performance attribute constitutes a large part of Apple's identity. Thus, keeping with tradition, the Powermac G5 must be unequivocally faster than any Pentium 4 based Wintels. Only a new performance gap that puts Wintels on the butt end of the stick will trigger the sort of mass migration to OS X that Apple needs for vitality. As Steve Jobs would say, we need Powermacs that are so fast they "make us shit in our pants".
Here's to hoping that the G5 dominates Wintels right out of the gate, so when Apple unleashes the Dual 3 GHz Powermac G5 upon the masses it will so totally destroy Wintels that Arstechnica will be forced to pull the plug on their Battlefront forum for wont of anything to discuss.
Apple, don't fsck up!
G5 + optimised soft for the G5 + panther = kick ass computer.
Originally posted by michaelb
Someone fill me in...
Are After Effects and Shake competing products?
Which is better?
Yes, and no...
After Effects is more suited towards motion graphics...
Shake is definitely a compositing tool...
There is some overlap in what they can do, and for the longest time After Effects was the standard in lower-priced desktop guerilla compositing... As with most Adobe products, plug-ins make to difference at the end of the day...
But Shake is much more suited for pure compositing work, stuff where you already have all of the elements needed for the shot, you just need to comp them together...
Plus, it is an Apple product...
Right now Adobe (and the boutique shop 'the Orphanage'...) is on my shitte list after releasing that ad dumping on Apple computers in LOTS of trade magazines... Bastards!
Just wern't right, Man!
;^p
Originally posted by MacRonin
...Right now Adobe (and the boutique shop 'the Orphanage'...) is on my shitte list after releasing that ad dumping on Apple computers in LOTS of trade magazines... Bastards!
Just wern't right, Man!
I just subscribe to DV magazine for the heck of it (I'm a DV hobbyist), and the first time I open my first issue, I see that fricken' ad. Can't wait for a new AE bake-off with a dual G5.
???????????????????????
We'll see how good AE 6.0 is... if they run another orphanage type ad for 6.0 you can bet Apple will fix the situation.
http://www.digitalvideoediting.com/c...es/reviews.htm
and adjust the times of the mac by multiplying the Dual 1.4 G4 system times by .56 ( 56% of the time recorded by Mr. White) you get a whole different impression of the Dual G4's speed.
minutes: seconds, winner in boldface type
Apple Power Mac G4 Dual 1.25GHz with 1GB DDR RAM
$3,589
Dell Precision Workstation 350
Intel P4 3.06 GHz,
1GB PC1066 RDRAM
$2964
(Estimated times in parenthesis)
1. After Effects: Simple Animation
:14 (:07.84 sec)
:07
2. After Effects: Video Composite
1:25(~50.4 Sec)
:54
3. After Effects: Data Project
3:47 ( ~2:10)
2:05
4. After Effects: Gambler
:43 (24.08)
:29
5. After Effects: Source Shapes
7:06 ( ~4:00)
4:14
6. After Effects: Virtual Set
8:15 (~4:40)
4:24
That would make the Dual G4 Stand up quite nicely.
also, when you thru the g5 in for $2999.
oh yeah baby!!
nice.
Using the Night Flight benchmark on the unoptimized After Effects 5.5, the G5 was 39% faster. Performance tests included Photoshop and Cubase, and a technology demonstration of the G5 masking video.
kiss my shiney metal ass, orphanage....
guess this mean that's abobe may be putting the extra effort forth to "optimize" their apps for the mac.