WWDC: New G5 officially announced, plus other goodies

[beaker]

Pixlets:
Jobs shows new QuickTime codec, Pixlets: a "studio-grade quality on a personal computer." It was requested by Pixar ;) It is very high quality designed for film quality material. Pixar + Wavelet = Pixlet. It features 48-bits per pixel source data and is the only codec to go over 24-bits. It has no inter-frame compression and offers easy editing, and back-and-forth scrubbing for the Pro market. HD/2 resolution is 960x540. Full screen, 24fps and stereo sound can be decoded on a 1GHz G4 Mac (TiBook). He shows high-quality Finding Nemo trailer in this quality with hi-definition sound. He also demo Rhe Matrix. Neither have visual artifacts.

Xcode:
Though this sounds cool, I am sure someone will say "aw thats been done before":

Jobs then talks about xcode, Apple's new developer tools, which supports distributed building, GCC 3.3 compiler, and much faster performance. It offers zerolink, predictive compile (compile as you write for about half the time), and "fix and continue."

[agreenster]

Makes me wonder if Pixar will be switching to the new workstations anytime soon.....

[Pyro2301]

Beaker:
Ars has a live update thing too...I started with MacNN but I find the Ars one is updated more frequently:

http://arstechnica.com/wankerdesk/3q02/wwdc-622.html

You have to manually hit refresh.

-Victor

[beaker]

This line on the ars page makes me really happy:

"Renderman benchmark: G5 is the fastest desktop in the world"

But he never announced renderman.

[MacRonin]

Yeah, i saw that to, and got real excited...

But nothing yet...

Maybe by the time Siggraph rolls around...

;^p

[Pyro2301]

Originally posted by beaker
This line on the ars page makes me really happy:

"Renderman benchmark: G5 is the fastest desktop in the world"

But he never announced renderman.


Hehehe ya I did a double take when I saw that...not quite sure if they ment what we all think they did.

[Pyro2301]

It has arrived:

http://www.apple.com/

[beaker]

The design is simple and looks really nice inside, but looks kindof ugly. I will have to see it up close.

[policarpo]

Man oh man....these machines sound like true masterpieces!

from the video i'd also presume that Renderman will be ported to OSX as well...I mean, if Ed Catmull says the G5 is the fastest on the planet...then man alive....why not have renderman on OSX!!!

it was also interesting to see the Luxology app...i wonder why they chose low-res meshes to spin around....but it was nice to finally see it. :)

I've been away from the Mac for 6 years, it's going to be great to be back on the platform!!!

Woo HOO!:bowdown:

[greekdish]

I just hope these machines really kick ass when released, and everyone benchmarks excellent results.....Im really tired of hearing how slow our Macs are...not that I use Macs purely for speed....but now, there wont be any criticizing anymore, and maybe people will just use whatever platform they prefer, as opposed to saying "Mine is faster than yours"....now, it should just be about "I like Mac OS better than XP", or vice versa (how it should be). Peace around the world has come!!!! :applause:

[StefanA]

I hope they can keep up with the development after it's released. Let's hope they wont go to sleep again. I would also like to see some other tests than those SPECS and 'photoshop' test. How about some Renderman and Mental Ray tests... that would be nice.

Anyhow, it was about time they released it. I'm glad for all the mac users out there that has kept up the faith.

.stefan

[beaker]

Originally posted by StefanA
I hope they can keep up with the development after it's released. Let's hope they wont go to sleep again. I would also like to see some other tests than those SPECS and 'photoshop' test. How about some Renderman and Mental Ray tests... that would be nice.

Anyhow, it was about time they released it. I'm glad for all the mac users out there that has kept up the faith.

.stefan
"Apple and IBM announce 3GHz G5 within next 12 months."

Atleast they are guaranteed for the next year :) 2.5 ghz in january and 3 next summer.

[policarpo]

i think the Apple and IBM alliance will be a good thing. IBM has been at it for decades, and Apple shows no signs of slowing down. We have a company known for innovation and a company known for breakthroughs working together to advance personal computing.

I think August will bring many surprises to all of us. Now, it's just a question of who will get to port the first native 64-bit 3D application to the G5!! Seems like Luxology might be ahead of the curve. :)

Also, does anyone know how we could convince Eyeon to port Digital Fusion to OSX so i'd be able to totally get rid of my PC's and protect my investment. :)

All you people going to Siggraph, be sure to take your 3D application of choice, your dongle, and a prebuilt scene so you can see how fast these puppies really are. :)

[Amyd]

Originally posted by policarpo
Also, does anyone know how we could convince Eyeon to port Digital Fusion to OSX so i'd be able to totally get rid of my PC's and protect my investment. :)

No WAY, dude. You want Apple buying off Eyeon from Leitch and next thing you know, all us Windows Fusioneers are left out in the cold with no upgrades? :eek:

Definately looking forward to some serious benchmarks of the new G5 systems, as the official Apple ones were always outright lies (though this time they look a tad more serious and scientific).

Hopefully all that means more serious competition for Intel and AMD... and hey, they even look prettier now, in a BOXX-sort of way. :)

Cheers,
Andrei

P.S.: I liked the Adobe PR quote: “Thanks to the hard work of Adobe’s engineers, Photoshop performs twice as fast on Power Mac G5s, when compared to any other system we've seen from Apple,” said Greg Gilley, vice president, Graphics Applications Development at Adobe."

[Per-Anders]

This stuff looks really sweet. I'm pleased for apple. I'd like to see some cinebench results showing the current top of the line systems on a face off. OGL has long been a problem on macs, so i'd like to see if the new systems are any better on that score, or if we will have to wait for x.3 before the ogl is upgraded to include such basic things as dual-planes. x3 with support for cards like the wildcat and full support for the 64 bit system will give these systems a bit of a boost from how they behave with osx.2.5.

anyhow, glad apple appear to be back in the ring (about bloody time too).

[Peter Reynolds]

Good news for anyone who enjoys mac shake

[Rhs_CG]

Makes me wanna chuckle in a Homer Simpson kinda way. Heh heh, yeah......
:beer:

[deepinspace]

Did you guys see the new 'luxology' program in the G5 introvideo!!!

[Kabab]

Hopefully this will push Intel/Amd to give us some more cpu's..

Goto love competition.

[passwr]

G5 specs exaggerated?

http://www.haxial.com/spls-soapbox/apple-powermac-G5/

[greekdish]

Originally posted by passwr
G5 specs exaggerated?

I dont think so...if you read the actual full report, you see that its even worse for the Intel hardware as its Red Hat Linux Pro running them, not the bloated Windows XP Pro. :shrug:

I really dont think Apple, after being so far back for a while, would boast these claims as to the power of the G5, and end up with mud in their eye if they arent substantiated. They can show only the facts that they choose, but they arent making anything up. Its called marketing, and of course they will highlight the best test results. :hmm:

[Lockstar]

an awful lot of buzz about this morning.

fake benchmarks this, floating point that, apple specs false, marketing hype etc etc etc etc.

will it never end.

All I know is that when watching the keynote, I saw the G5 perform a 100 PS actions on a 360Mb twice as fast as a dual Xeon. Run Mathematica twice as fast, and logic playing a 1000 voices simultaneously, utilizing 25% CPU. Oh and Ed Catmul says it performs the fastest Renderman benchmark. and Alias are apparently presenting at the mac show in NY. (hopefully Maya optimised)

These are realworld apps that I use. And now i'll be able to get my work done faster (is this not more important than GCC compiler issues?)

Apple have released an impressive machine, congratulations. it will not overtake the world, and no doubt Intel, AMD, Sun etc will provide a response. but it's good that the playing field has been levelled and the Motorolla bottleneck has been eradicated.

stop the whining, stop the bigotry and stop the zealotry. :)

[Amyd]

With the important emphasis on waiting to see the same kind of real-world applications tested by independent sources on a G5, because it really is very easy to both choose a specific set of actions/steps or purposely misconfigure one of the test systems to make it look like it is 2,3,10 times slower (which unfortunately Apple has been known to do, at least in the past when it comes to... Photoshop benchmarks, for instance).

Have they in the pasted released samples of systems to hardware test sites before the shipping date like the PC companies do from time to time or will we have to wait until August/September to see these kinds of tests and benchmarks? Can't wait... :D

Cheers,
Andrei

[thms__0]

i love the Powermac G-5 the perfect workstationhttp://www.cgtalk.com/newreply.php?s=&action=newreply&threadid=71379#and it looks goodhttp://www.cgtalk.com/newreply.php?s=&action=newreply&threadid=71379#

[DaveW]

Like Amyd said, Apple is known for fudging benchmarks to make their systems look better. I don't doubt that the dual 2ghz Mac is faster than a dual 3ghz Xeon for some things, but it is not 2.5 times faster. The SPEC scores alone show that something is wrong; lots of third parties have done SPEC tests and show x86 systems have much higher scores. We also know that for some of the tests they disabled hyperthreading on the Xeons with no explanation (obviously because even with the tests slanted in Apples favor, with ht on the Xeon still would have won). Ed Catmull may be right about Renderman being faster, but he doesn't say it's 2x faster like Apples bogus benchmarks would have you believe. When these systems actually ship we'll get the real results and I can guarantee you the speed will be much closer to current x86 speeds. It's still a badass system and it's nice to finally see Apple have some competitive hardware again, but Steve Jobs is the king of hype and bs, so take his outrageous claims with a huge grain of salt.

[Amyd]

There are other very dubious claims in the performance PDF available on Apple's site.

For instance, they list very low scores in Quake III Arena at 1024*768 for the Pentium IV platform:

G5 with 9800 Pro: 337 fps
Dell 8300 3 Ghz : 275 fps

while a casual stroll around the hardware benchmarking sites reveal the following results for a 3 GHz Pentium IV system with a 9700 Pro (!) in the same test conditions:

Anandtech 3Ghz: 394 fps

That's more than 100 frames per second (not taking into acount the fact that the Radeon 9800 Pro would score 10-20% more on the Anandtech rig) shaved off the Dell system for some unexplainable (to me at least) reason... :shrug:

And I don't even want to go into the Serial ATA page, where Apple doesn't even bother to compare apples to apples and presents graphs of Serial ATA on Mac vs. Parallel ATA on PC ?!?

Cheers,
Andrei

[My Fault]

Originally posted by Amyd
while a casual stroll around the hardware benchmarking sites reveal the following results for a 3 GHz Pentium IV system with a 9700 Pro (!) in the same test conditions:

Anandtech 3Ghz: 394 fps


I'm looking at Anandtech right now and for the same conditions the top Intel Pentium 4 3.00GHz (875P-800MHz) at 1024x768 scores 213.8 fps (http://www.anandtech.com/cpu/showdoc.html?i=1810&p=7) . Are you sure the 394 fps isn't for 640x480? Where is the link to that 394 fps claim you made?

[Amyd]

Originally posted by My Fault
I'm looking at Anandtech right now and for the same conditions the top Intel Pentium 4 3.00GHz (875P-800MHz) at 1024x768 scores 213.8 fps (http://www.anandtech.com/cpu/showdoc.html?i=1810&p=7) . Are you sure the 394 fps isn't for 640x480? Where is the link to that 394 fps claim you made?

That is an Unreal Tournament 2003 test, not the Quake III Arena demo 4 that Apple used in the PDF claim.

You can find the Anandtech results here (http://www.anandtech.com/cpu/showdoc.html?i=1834&p=10) (sorry for not thinking about linking it in the original message).

Cheers,
Andrei

[My Fault]

Originally posted by Amyd
That is an Unreal Tournament 2003 test, not the Quake III Arena demo 4 that Apple used in the PDF claim.

You can find the Anandtech results here (http://www.anandtech.com/cpu/showdoc.html?i=1834&p=10) (sorry for not thinking about linking it in the original message).

Cheers,
Andrei

Woops, I have been playing too much UT2003 lately to have read Quake III as UT2003 :)

Yowch, that is so far off it is not funny. Bad Apple! Hopefully we'll ge to see some independant tests in the next few weeks and see how they really compare.

[greekdish]

Originally posted by Amyd
There are other very dubious claims in the performance PDF available on Apple's site.

For instance, they list very low scores in Quake III Arena at 1024*768 for the Pentium IV platform:

G5 with 9800 Pro: 337 fps
Dell 8300 3 Ghz : 275 fps

while a casual stroll around the hardware benchmarking sites reveal the following results for a 3 GHz Pentium IV system with a 9700 Pro (!) in the same test conditions:

Anandtech 3Ghz: 394 fps

That's more than 100 frames per second (not taking into acount the fact that the Radeon 9800 Pro would score 10-20% more on the Anandtech rig) shaved off the Dell system for some unexplainable (to me at least) reason... :shrug:

And I don't even want to go into the Serial ATA page, where Apple doesn't even bother to compare apples to apples and presents graphs of Serial ATA on Mac vs. Parallel ATA on PC ?!?

Cheers,
Andrei

Well first off...those are two different versions of Quake....Anandtech is NOT running the latest version of Quake, which is 1.32 that APple is using....why is Anandtech using 1.29f?? Its some sort of tweaked version maybe?? Also....Anandtech can be using an entire different config than what Apple is using...Apple can just be doing default max res at 1024, whereas most of these game sites love to tweak to show as many fps as possible. I for one NEVER get over 300 fps in Quake with my 3 Ghz Sony and GeForce Ti 4400. Ive NEVER been able to come close to any of the "benchmark fps" on gaming websites. Sorry, im not a techno geek tweaking my settings all the time. Im happy with a smooth 150 fps any day. :D

[Sieb]

Thats why you never trust performance claims by the home team. What company is going to publish marketing material that pits their machine just as or less equal (negatively) to the compitition. Not to mention that at the root, the 970 is still a different platform from x86 (or 64bit). Some of the tests didn't even have hyperthreading running on the PIVs.

Always wait till the third partys (preferebly no mac only) to run tests and then average all their results (between the different parties).

[Amyd]

Originally posted by greekdish
Well first off...those are two different versions of Quake....Anandtech is NOT running the latest version of Quake, which is 1.32 that APple is using....why is Anandtech using 1.29f?? Its some sort of tweaked version maybe?? Also....Anandtech can be using an entire different config than what Apple is using...Apple can just be doing default max res at 1024, whereas most of these game sites love to tweak to show as many fps as possible.

O'course they are using an entirely different config, it's a PC. :)

Seriously now, if you look at the Anandtech page I linked it provides the same kind of explanation that Apple provides in the PDF:

Anandtech:

"[...]We used our old 1.29f build of the game with the classic demo "four" at High Quality defaults, with everything maxed out at 1024x768.[...]"

Apple:

"[...]We tested Quake III Arena 1.32 at two resolutions to differentiate between demands on the processor and demands on the graphics card. The test was conducted with all
default settings using the Demo_4 benchmark.[...]"

Except for the difference in the Quake III version, the settings seem to be pretty much the same (one could suspect Apple of using lower quality settings, in fact, which I hope is not the case). And I can't see in the release notes (http://www.idsoftware.com/downloads/132_changes.html) of the game update any reason for that >100 fps difference. Usually Anandtech (from what I know) are pretty serious when it comes to testing and benchmarking.

But of course, it could be that the slight difference of versions and settings could account for that huge gap... which is why I said the Apple results were "very dubious claims" and not outright lies, until some independent resources can verify them. :)

Cheers,
Andrei

[greekdish]

O'course they are using an entirely different config, it's a PC

:p LOL....thats not what I meant by config. I meant the settings config that is used in games with the Quake engine such as Quake 3 Arena and Wolfenstein, Jedi Knight II, MOHAA, etc. Anyone can go in and tweak them to your hearts content. For instance, you can still have high quality settings, but tweak lifetime of marks, particles, AA or not, hardware or software fog, etc. While it does seem like Anandtech can tweak the bejesus out of those tweaks as opposed to default high quality settings, Im sure for Apple, the configs were the same on both Macs and PCs....default. Anandtech also probably used the same config. Its not like they are fudging the results, they just used the config they have. Ive seen many different websites with varying degrees of framerates. Anandtech seems to have the highest framerates anywhere by a wide margin. Ive personally never seen anything remotely close to their results, and I have a 3 Ghz P4 with a GeForce 4 Ti4400, not exactly a nightcrawler there. :drool:

[Amyd]

That PC remark was a joke, I understood what you meant in the first place. :D

The thing is, since both sources say they used default configurations without mentioning any tweaks, I am inclined to assume that in fact they had almost the same settings and so the comparison stands. Plus those defaults have in fact been... defaults among the PC benchmark sites for quite some time, so my guess Apple followed the same trend.

Anandtech seems to have the highest framerates anywhere by a wide margin.

That is absolutely not true. It is very easy nowadays to do a survey since world+dog are doing benchmarks on the newly released P4 3200 and most of them include Quake III Arena benchmarks and the funny thing is they have almost the same hardware configuration between them (Canterwood, Radeon 9700 Pro, etc.). Here is a selection for the PIV 3Ghz:

X-bit Labs: (http://www.xbitlabs.com/articles/cpu/display/p4-3200_5.html) 405 fps
Tom's Hardware: (http://www6.tomshardware.com/cpu/20030623/p4_3200-07.html) 402 fps*
Hexus: (http://www.hexus.net/review.php?review=574&page=7) 390 fps
Sharkey Extreme: (http://www.sharkyextreme.com/hardware/cpu/article.php/3261_2226901__5) 387 fps

*For the sake of accuracy, THG uses a different demo for the benchmarks so therein might be a source for differences

As you can see, Anandtech's scores fall right in the middle and anyway the difference between the various sites is nowhere near the 100+ fps that are substracted from the Dell score in the Apple benchmark.

To end the sort of offtopic we have here, I'll remark that on my 2.6C/875P (Gigabyte)/Radeon 9700 Pro config I usually seem to get in benchmarks (including Quake III) within around 5-10% of the published benchmarks in the hardware review sites. So... :thumbsup:

Cheers,
Andrei

[CGTalk Moderation]

This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.