Slashdot Mirror
Apple Acquires Silicon Grail
mac writes "Silicon Grail's web site has an interesting update: it has been acquired by Apple. Their product RAYZ and Nothing Real's Shake are the two major products, as far as compositing software goes. Nothing Real was bought by Apple also back in February. With both companies held by Apple, who will fill the void in the Windows and Linux?"
What... is your name!? Steve Jobs.
What... is your quest!? To buy the Silicon Grail.
What... is the average performance of a Dual G4 when compared to an Intel Pentium 4? What, Rambus, or DDR powered?
I don't know that... AAAAAAAAARGHHH!H
Here's what it is.
Better writeups, please.
I have been pwned because my
There are no shortage of video editors out there and a quick search of Sourceforge for "video editing" shows a good chunk of projects rolling along.
No Zen is good zen
Yep. Dude, it's redundant to demand better writeups. You've been here how long?
Or am I being ridiculously optimistic? Do I need to Think Different to understand Apple's financial decisions?
Here's what I think will happen :
Apples not going to cut of a revenue stream, so clearly they'll keep Windows / Linix / other versions. Its in their best interests to allow these acquisitions to generate revenue, if for no other reason than to mititgate acquisition costs.
Intermediate term, I wouldn't be surprised to see "enhanced" versions on OS X only, followed by a longer term retirement of Windows / Linux / other versions.
This way they entice folks over to their high margin platform (Mac), while not pissing them off and giving them adequate time to retire the older hardware gracefully.
While it makes sense for Apple to buy these packages, it may not make sense to immediately assume the company will make them MacOS-only in either the long or short-term. There are a number of Apple technologies including AppleWorks (Formely ClarisWorks) and FileMaker Pro which continue to have Windows versions produced to this day. Rather, these purchases simply let Apple showcase the advantages of the MacOS X platform by *forcing* a port of these products to MacOS X and making sure that port takes the fullest advantage of the MacOS X toolsets.
At least that is my take.
I read that as "composting software", and was trying to figure out why Apple people would buy special software for recycling waste files into free space.
Actually, this is cool for Apple, because years ago you could refute those who said Apple was the graphics platform by pointing out the lack of CGI and high end compositing software for the platform (Premiere isn't high end before someone says that), but it appears Apple is attempting to remedy it.
The real question is, how long after they force developers to poorly port things will they sell it to Avid? ;)
It all is starting to work out now.
;)
They already have the best video editor in Final Cut Pro. Then they released Cinema Tools to help convert between film and video, easing the editing process for cinema features. Then came the Xserve, paving the way for the server / heavy workstation in the creative business. Now by combining the best of the two leading technologies in the high-end compositing market Apple can take over the entire movie business by simply being the best choice.
Sneaky, but I like it
"I tend to think of OS X as Linux with QA and Taste", James Gosling, creator of Java
I would assume that the people who've bought this software for Windows consider it a big deal. For the most part, people generallydon't just port commercial software to a platform just because they feel like it. They do it because they've done or recieved some kind of market research that tells them that people will actually buy it.
Not that there was much of a void. AfterEffects on the low end and Discreet products for the high end.
http://www.discreet.com/
From an earlier article on the fate of Shake :
``Apple has declared that Irix and Linux versions will be developed at least through 2003.''
To me that suggests that there is no Linux void yet. Also, the fact that they say they will keep developing for those Unices but not for MS Windows suggests that perhaps they will Go Unix with this, which they can do thanks to Mac OS X's Unix roots (kudos to NeXT for coming up with this brilliant idea).
---
Economics is extremely useful as a form of employment for economists.
-- John Kenneth Galbraith
Please correct me if I got my facts wrong.
Now, what this effectively does is make it so that the people who used to shell out big bucks for the software product now have two choices: find a new piece of software, or spend less money than they priviously wold have spent on an upgrade, and buy a powermac and the mac version of the software.
This is actually quite a brilliant strategy. Think of it this way: I use product x. Product x costs $20,000 and an upgrade costs $5,000. Product x is the core of my business. I use windows PCs to run product x. Now apple buys company x, who makes product x. Nothing changes for a few years. Product x's windows support is phased out, and the mac version's price is dropped to $2,000 for a new product, and $999 for an upgrade. I can now purchase a powermac for $5,000 and a product x upgrade to mac for $999 and end up spending only slightly more than I would have otherwise. Furthermore, in the future, upgrades will be very very cheap. Or, if I don't like apple, I can stop using product x, and instead use product y, which, since I never used it before, now costs $20,000, and has an entirely different interface...
See why Apple's strategy is smart?
Let's not think too inside the box about Nothing Real and Silicon Grail. Apple's also aggressively going into the MPEG-4 market. Now, MPEG-4 has a wonderful feature, foreground and background separation. As I understand it, you can send the weather map once, and then send just the weatherman moving in front of it from then on. As they'd say up here in New England, you get wicked-good compression that way.
What's the killer application for automated background separation? Video-conferencing, of course, or what Apple might call iTalk. Video-conferencing has not been well received, largely because of bandwidth problems. MPEG-4 gives you really nice full-frame compression, but add in the automated layer separation, and it gets way better. It might even be good enough to do on a GSM phone. Cable modems are definitely more than good enough.
So, who has the technology for separating people from their backgrounds? Hollywood, of course - that's what they use for putting live actors into special effects. Who's considered the best by Hollywood? Nothing Real and Silicon Grail, of course.
So, Apple builds this into the January version of OSX and shows an ad with the couple who got married in Hawaii last year; they've got a kid now, and Grandma gets to watch him take his first steps live because she's got an iMac that's on the cable modem 24/7. Digital lifestyle.
I expect the QuickTime team are the guys waving the landing lights for the Nothing Real and Silicon Grail tech. Even if I'm wrong about the application, there's no better place in Apple to absorb the technology.
There's probably more going on here besides just beefing up up Final Cut Pro.
My God, it's Full of Source!
OUTSIDE_IP=$(dig +short my.ip @outsideip.net)
...that Silicon Grail uses a whole bunch of code that they licensed from Kodak's Cineon product, and which Grail has no transfer rights to? Knowing Kodak, they won't do anything about it, but it would be interesting if they withdrew the license entirely and left Apple holding the bag on a much less interesting product.
Of course, the guy three cubes over who is porting other bits of Cineon code to Linux would probably be out of a job if that happened, so I hope it doesn't happen.
The next Cmdr Taco duplicate will be ready soon, but subscribers can beat the rush and see it early!
Why on earth would Apple not want to immediately integrate its new video editing holdings into Quicktime as completely as possible as immediately as possible?
Rayz, Chalice, and Shake are not editing programs. They're compositors. Big, big difference.
Nobody needs QuickTime for a compositor, except possibly to support reading and writing of image formats. And Shake, Chalice, and Rayz already have those problems solved.
(At this point, you may almost almost call Quicktime a full fledged operating system. The thing is such a beast that i have heard it said that Apple does not actually port Quicktime to Windows-- that instead, they wrote a limited Wine-like "emulation" layer translating all the parts of the Macintosh APIs that Quicktime uses into the equivilent parts of the Windows APIs)
Are you just making stuff up at this point, or what?
Are you sure about the lack of transfer rights? Do you know in detail how it was worded? I'm pretty sure that there is next to no Cineon code in Rayz, just the motion interpolation code.
-- SIGFPE
This was received from an anonymous source...
Now everyone is going "ooooo, what does that mean for the rest of us"? Meanwhile, my ex-friend Bill, buys up companies left and right and there is barely a ripple in the Linux and Apple community. So why do I seem more threatening?
Personally, I think it is my turtle-necks. See, I learned from Grace Slick that nothing hides the look of age than a turtle-neck. Especially black because it a) looks cool b) hides that extra "executive weight". But it threatens people that I can look cool AND youngish at the same time. Bill looks like someone's Grandfather - or Mr. Burns from the "Simpsons". "Smithers, buy up Freedonia" - see how that would just seem natural coming out of his mouth?
But I buy two companies and BOOM I'm killing babies and eating their entrails.
My point is that first, Apple is a business and as a business it attempts to stay in the black, much like my turtlenecks. Second, get a grip. Mergers happen all the time. Some are good and some are bad. Third, I'm still cool, right?
Your Pal,
Steve
IANAL, but I've seen actors play them on TV
Why not follow the same couple/pros doing their video editing on a Mac throughout their entire marketing campaign?
Jon and Jen use Final Cut Pro and DVD Studio to make their DVDs for clients. They use MPEG4 to stream content using an approach like you outlined here. Then the couple's kid uses iMovie to make presentations for her class project.
Not as obnoxious as Dell's Steven, you identify with real users and can target specific markets for ads. Heck, you could even have XServe in there somewhere.
I'd prefer replies to Karma. I have plenty of Karma.
Digital Fusion has, IMO, a better interface than Combustion or Shake. I only used RayZ for about 30 minutes but I didn't like it much. I like to be able to drag the viewports to the places I want, instead of being stuck with the layout that the authors decided I should use.
RayZ did seem to do more stuff than the other programs, but I really can't speak about quality because I didn't use it for long enough.
RMN
~~~
#include "ShamelessPlug.h"
More info on Digital Fusion here: www.eyeonline.com
Why would anyone engrave "Elbereth"?
Though, truth be told, Discreet did recently shut down their entire Combustion development office...
Why would anyone engrave "Elbereth"?
"Apple has acquired technologies from Silicon Grail"
They don't list the technologies, but they do include RAYZ and Chalice. They don't mention the Cineon technology.
Silicon Grail is now an empty company. What's Ray going to do next? And what does happen to the Cineon stuff?
Why would anyone engrave "Elbereth"?
Everyone would then be on an even playing field.
But...jeez...who has time to waste on Linux.
** Please note that the following is purest, purest speculation. ** ...
It is my personal theory that one of the big reasons we will not be seeing a linux version of quicktime is because
** Please note that the following is pure fact. **
There are only two reasons that you will not be seeing a Linux version of QuickTime.
A) No one has come up with a revenue model to justify Apple's investment in it. And with no significant portion of either production or consumption of video taking place on Linux desktops, I sincerely doubt anyone ever will.
B) No outside party has offered to provide Apple with the necessary investment. Apple would be more than happy to do so at anyone at all's request (well, the QuickTime team anyway, can't speak for top management but I doubt they'd be offside) it's just that the Red Hats etc. haven't put the dosh up.
END. OF. STORY.
How would they hope to make enough profit on this arrangement to get their money back? There just aren't enough people doing compositing out there to do it for marketshare alone.
Now, if they were to buy up 3D companies, THAT market is a lot bigger...
Why would anyone engrave "Elbereth"?
"Who will fil the void for Windows and Linux?"
Lets see...windows...how about Combustion?
and Linux.....the stuff put out by Alias|Wavefront (cant remember the name right now)
I'm out of my mind right now, but feel free to leave a message.....
You really should try to understand the difference between "trolling" and "fact". A fact isn't necessarily something you like, it's just something that's true.
/. "culture"...
The G4 is slower than the Intel / AMD alternatives (see this test, for example), and PCs are cheaper than Macs (especially if you run Linux and thus save the "Windows tax").
Once you're inside a compositing (or animation) program, the actual operating system isn't relevant. What matters is quality and speed. If a program exists for two platforms, then its quality will be the same. So it boils down to speed (or, more precisely, the price / speed ratio, because all these programs can be set up in render farms). So the platform that gives you more "bang for the buck" will inevitably win.
Now, Apple could gain market share by killing the competition, but there's no way they can kill all the competition. Discreet rules the high-end and there's no way Apple can buy them (Autodesk is far too big). So to be competitive, Apple must either sell very cheap render nodes (the Xserve could be a step in that direction) or come up with much faster hardware (wasn't the G5 supposed to be out yet?).
Unlike 99% of people writing in this thread, I actually do work in animation and post-production, so I have some clues as to what I'm talking about. Maybe that just doesn't fit in with
Will I go from "troll" to "insightful" if I say Bill Gates is the devil...? Personally, I think those posts should be modded as redundant. Everbody knows it already.
RMN
~~~
"Separating people from their backgrounds" is called "keying". Shake has no keying technology of its own, they licenced Primatte from Photron. I don't recall what RAYZ/Chalice used, but certainly keying was not the focus of that product.
If they wanted keying technology, they'd buy Photron themselves, or Ultimatte or Zbig (too late) or maybe even a compositor with its own keyer.
Why would anyone engrave "Elbereth"?
Apple purchased Nothing Real in February 2002
But I thought they bought that company in February what does it mean they purchased nothing? Oh wait the name of the company was Nothing Real? Ohh!
I stole this Sig
Actually I was thinking more about AMD than Intel. As someone said, the P4 lets programs run faster if you optimise the code for SSE2. The Athlon XP made existing code run about 15% faster without having to re-write anything.
The Athlon's FPU is much, much faster than the P4's.
And optimising for SSE2 or Altivec will only take you so far. 3D Studio MAX has recently been optimised for the Pentium 4 (by Intel programmers) and it still runs faster on Athlons. Some algorithms are simply not suited to vector optimisations.
Apple may be moving to DDR266, but x86 platforms are moving to DDR333 and DDR400. And the Athlon also benefits a lot from faster memory (the P4 has RAMBUS, wich is even faster). Anyway, 3D rendering for example doesn't depend too much on memory performance; it depends on pure FPU power.
If the Xserve was cheaper and / or faster, I could see it as an alternative to dual Athlons / Xeons. I guess it'll depend on the performance of the G5 and how it compares to the x86 alternatives (the Hammer seems to be coming along nicely, and the P4 can probably be pushed up to 3 / 4 GHz during the next 12 months).
RMN
~~~
From an article located at creativemac.com (from the same publishers as digitalvideoediting.com):
Maybe that's why Apple made this purchase, so that they can become a source for compositing software that takes advantage the G4's attributes.Sorry, but your original message reads like a troll. It has little to do with the subject at hand, yet takes every opportunity to bash the overall subject matter of this category.
NetInfo connection failed for server 127.0.0.1/local
The main conjecture here is that Apple is buying these companies to integrate their technology into QuickTime for future Digital Lifestyle applications, not to resell their existing products. I suspect the MPEG-4 layering features are relevant because that's the technology they're focused on that would most benefit from really good compositing.
;)
You make a fair point, but the ownership of the keying technology does not significantly change the probability of this outcome. It may even make more sense, since QuickTime is primarily an output platform - they might decide to do the keying in a hardware device (camera) with guts similar to an iPod.
Seeing as the aquired companies had existing contracts with the keying companies your mentioned, and Apple bought them, the contracts should carry over. If the contracts were sufficiently elastic it may not make sense for Apple buy them. If they were not good enough, your list may make a good candidate list for future aquisitions. I'll check back when they announce next-quarter's acquisition.
(being June 13th, this probably also signals that Apple will show a nice profit for the 2nd quarter, such that they can afford an aquisition).
My God, it's Full of Source!
OUTSIDE_IP=$(dig +short my.ip @outsideip.net)
Of course he means QuickTime 2, not Windows 2.0.
Nope, you've got it backwards. QuickTime on OS9 was an extension, which you could readily disable and remove, and as such no part of the standard MacOS toolbox relied upon it. (QT on OSX is a little more integrated but the historical separation remains, QT is built on top of the toolbox, not vice versa.)
QuickTime uses the Mac toolbox fairly heavily -- it's not about media compression, it's about time-based playback of media. Compressing and decompressing is just something you -sometimes- need to do as part of this. A lot of the code is getting it from the uncompressed format to the screen/speakers in a timely manner. It uses windows, threads, events, timers, component manager, resource manager, file manager, sound manager, just to name a few. As some other folks said, they basically ported a good chunk of the MacOS Toolbox to Windows to make it work.
I won't argue too much with that one. I should point out that one of the reasons they didn't port it was because as big as it might seem to some young'uns
They ported the Mac toolbox to Darwin first - aka Carbon. Then, and only then, they brought over QuickTime, building it on top of Carbon. You'd have to port a large part of Carbon to Linux. Mind you, it would be possible, but there probably isn't a large perceived ROI for the time involved.
I'm sure Apple wouldn't mind extending their media-player dominance to another consumer desktop platform, but Linux (and *BSD) are not consumer desktop platforms yet. When binary distributions become more feasible and the userbase grows, that's when you might see QT for Linux.
(former Apple engineer)
So next you're going to say I can't compare a 1.8 GHz Athlon with a 2.53 GHz Pentium because the Athlon has a lower clock speed...?
You can only compare products that exist. After Effects is one of (if not the) most common compositing program in Windows and Mac platforms. The Dual G4 used in the test was the fastest Mac that existed at the time (it still is, I think). Also, both systems were similarly priced ($2564 for the PC, $2870 for the Mac). They even removed some of the PC's memory to make it match the G4.
It's about as fair as it gets. In fact, I think they should have used a faster PC (that did exist), even if it was a more expensive.
Now, I find it rather odd that Affter Effects (mady by Adobe, which is traditionally an Apple partner) is so poorly optimised for the Mac, and so well optimised for the PC. If this is the case, then people running AE on the Mac should probably drop Adobe a note.
Can you post a link to any other independent test (ie, not made by Apple or a PC manufacturer) where the same program is shown running on two top (PC & Mac) workstations? Say, Photoshop, 3D Studio MAX, Lightwave, Combustion, etc...?
That link you posted compares single G4s to dual G4s, but (strangely enough) not to dual Athlon MPs or Xeons.
And my message has everything to do with the subject. Selling cheap software my be important for the low-end market, but for the high-end (professional) market, what counts is deadlines. And that boils down to render speed. Why would anyone buy 50 G4 render nodes when they can get the same speed with 30 Athlon MPs or Xeons...? And if their budget is enough for 50 G4s, then they can probably buy 70 Athlons and finish the jobs even faster. Rendering faster doesn't just mean you can meet the deadlines; it often means you can do two versions of each project, and keep your clients happy.
RMN
~~~
Whether this is moral or not, Apple is basically doing a Microsoft. Buy up key companies whose technology forms the basis of certain sectors and starve the competition.
I predict they will do this with Maya as well, and if things go well, possibly Softimage.
This is a good insurance policy for them to stay alive. They have a similar position in the DTP market but eventually Quark is going to go the way of the dinosaurs and the market will be up for grabs. Most Adobe and Macromedia products are written for better integration on Windows these days with the Apple ports lacking somewhat in pollish. Apple should do more to ensure that it's niche in DTP is renewed as well.
Everything you said is wrong :).
QuickTime 2 was the first version for Windows. It was playback only - authoring required a Mac. It wasn't really a port, just a player library that could do QT files.
QuickTime 3 for Windows was a ground-up new version that supported Authoring. Since QuickTime for Mac had huge dependcies on the underlying MacOS "Toolbox" for QT3 for Windows they actually ported over a huge portion of the MacOS APIs so it could run. It was complete enough that Apple had to specifically request software vendors not use QuickTime as a Mac to Windows porting library. And some still did, like Media Cleaner Pro 4.
QuickTime is a whole media architecture. It does compression, sure, but lots of other stuff. It is a major enabling technology for video editing, and also does panoramas, audio playback, etcetera. Its complexity is on the same order of magnitude as the Linux kernel.
Apple doesn't get any money from QuickTime licenseing. While you need to license the installer from them, it is free as in beer. You just need to send them two copies of your disc for regression testing against new versions of QT.
QuickTime for MacOS X is Carbon, which means it uses the port of the Mac toolbox for MacOS X (in the same way it uses is own internal port on Windows). Porting it to Linux would require porting this as well. This is far from trivial - QuickTime needs to talk directly to low level hardware like sound cards, clocks, video cards, etcetera. This aren't things that are well unified under Linux. QuickTime is extremely heavily tested by a large testing team. So even if they did it, they'd have to pick a few Linux flavors to test against. The kinds of things QuickTime does are the kinds of things that break on random distributions.
I've heard that the Windows port took something over 100 engineer years, and I imagine Linux would take at least as many. That's, VERY rough ballpark, $20M.
Think Apple could see an additional $20M in net revenue from having a Linux port?
My video compression blog
An AC wrote:
> My guess is that Stevie is after two things...
>
> 1 - kudos for having snazzy companies that do
> 'cool things' and can do it on a Mac,
>
> 2 - the technology in the companies (the main
> reason), to integrate them into existing Apple
> products especially Final Cut Pro. eg: the
> Cinespeed code that Silicon Grail got from
> Kodak.
With #2 you are getting close. Let's look at what Apple has put together and/or invented themselves:
QuickTime
Digital Video with Firewire
A field video workstation: Ti Powerbook
A pro video/graphics workstation: G4 Power Macs
A high end video workstation/cluster server: XServe
OS X: child of Mac and NeXT, grandchild and heir to the name of UNIX
Final Cut Pro
Cinema Tools for Final Cut Pro
DVD Studio
Shake
"technologies" purchased from Silicon Grail
a CEO with experience from running Pixar
Also available for the Mac from 3rd parties:
Photoshop and other high end graphics software
The big names in 3D in Hollywood: Maya and LightWave
Get the picture yet? Apple plans on taking Hollywood by storm. Avid had better start thinking about how low their prices can go because next to them, Apple complete with hardware is a fraction of the cost.
How does this benefit us ordinary folks? Plenty. The New Apple dances to Mothra's songs. The more the movie and music industries come to depend on Apple, the louder Steve Jobs will preach, and the more they will listen. And what Steve Jobs is preaching is an end to the insanity the MPAA and RIAA are currently practicing. What? You thought Sony and Universal just decided to market burnable songs on the internet out of the blue? Or is it because of Jobs' speech at the Grammys?
The fate of Linux in Shake and the others probably depends on three things:
1) how well XServe does in the marketplace
2) how well Shake et al for Linux sell from now thru 2003
2) how fondly attached present customers of Shake et al are to Linux
So if you want to see Linux for these products in 2004 and you are a current or prospective customer, let Apple know now.
For those of you who want full QuickTime for Linux, work on your desktop marketshare and see if "Uncle" IBM will lend you the millions neccessary to license the codecs. Maybe then if you ask nicely, Apple will do a version for you. Lots of downloads of the QT streaming server for Linux would help convince Apple of your interest.
"Mothra's attack is working."
-- Shouta, "Mothra 3: King Ghidora Attacks"
Apple is being smart. A LOT of people out there were looking to Rayz as a Shake replacement now that NothingReal went to the dark side.
With Shake gone...... well if you want to work in compositing u need to start evaluting Discreet, or budget for some fruit colored plastic computers.
Keying is another matter. It's quite difficult to do good keying, even when the background is a nice, uniform, unique shade of colour, which it never is. You have noise, transparency, fringing and spillover to deal with. That's why Ultimatte etc devote their whole companies to the subject. Buying keying technology makes more sense, but you wouldn't pay many millions of dollars to buy a company that just licenced a keyer, you'd licence the keyer yourself, or you'd buy the keying company.
But to separate talent from background when there's no blue or green screen behind them is much harder still. You can do a difference key, if the background is completely still, but the results usually need manual cleanup. Otherwise, you have to rotoscope each frame - "cut out" the talent from the background by hand. It takes ages, but it's often the only way. Shake, Chalice, Digital Fusion etc, are good for doing this, but they don't do it for you, not by a long shot.
The "digital lifestyle" you describe has no relationship to this article. Not unless you want Granny to paint her computer room blue & build one of these into her iPod-camera, or hire a team of high-speed rotoscopers.
Why would anyone engrave "Elbereth"?
There is no Quicktime for Linux.
[ the following command shows packages available on a Gentoo source based installation. It is somewhat analogous to 'rpm -q -a |grep quicktime' on a Suse/Red Hat/Mandrake box, or 'dpkg -l |grep quicktime' on a Debian box]
# emerge -s quicktime
[ Results for search key : quicktime ]
[ Applications found : 2 ]
* media-libs/openquicktime
Latest version Available: 1.0
Latest version Installed: [ Not Installed ]
Homepage: http://openquicktime.sourceforge.net/
Description:
OpenQuicktime library for linux
* media-libs/quicktime4linux
Latest version Available: 1.5.5
Latest version Installed: 1.3
Homepage: http://heroinewarrior.com/quicktime.php3
Description:
quicktime library for linux
As you can see, there are at least two quicktime libraries available for GNU/Linux (mplayer will play quicktime videos, as will xine IIRC). In addition, codeweavers have a wine-based quicktime plugin project as well, so Linux support is covered pretty well.
Apple may not be bothered to produce a GNU/Linux version (hardly suprising since GNU/Linux relegated Apple to 3rd place ranking in PC OS marketshare), but free software developers are quietly going about supporting the format under GNU/Linux and *BSD, sans the usual press release fanfare that accompanies Microsoft and Apple products.
The Future of Human Evolution: Autonomy
Apparently you "do 3d", so you're concerns are valid even as your points *may* not be.
Meeting deadlines is important.
Nothing says a G4 cannot meet deadlines. Without Shake or Nothing Real, is there software available on the Mac that even allows you even consider Apple?
A 1GHz G4 vs a 1.6GHz AMD is *not* a hands down victory for the AMD any more than a 1.6GHz AMD vs a 2.4GHz P4 is a hands down victory for the P4; just on simple analysis, the G4 has a SIMD unit *twice* as wide as the AMD solution, and has 4 of these Altivec units to the Athlon's 3 3d!Now units.
So more than anything else, a G4 is memory starved, though the new XServes help a bit with their DDR chipset.
Just on a stupid upper limit, at the same clock a G4 can process over twice as much data, Altivec vs 3d!Now. If anything *both* of these processors are memory constrained.
So the G4 *already* keeps up with the AMD, and the AMD *already* keeps up with the P4, excepting the fact that there doesn't exist the software parity between the platforms... which Shake and RAYZ may help alleviate.
Yes, you have the advantage of experience in this arena, so I'm largely arguing *possibilities*, but the whole point of Apple buying Nothing Real and Silicon Grail is about possibilities and Apple's execution into reality.
Add on top of this that Apple does have Darwin->x86 render nodes, XServe render nodes, *and* G4 optimizations up their sleeves, and it's not as *hopeless* as you make it up to be...
GPL Deconstructed
The G4 is slower than the Intel / AMD alternatives (see this test [digitalvideoediting.com], for example)
It actually depends. On code that takes advantage of Altivec, you can see dramatically faster performance on the Mac. For instance, we have custom code here running on a dual 800 Mhz Mac that blows the doors off of an Intel P4 at 2.2Ghz.
and PCs are cheaper than Macs (especially if you run Linux and thus save the "Windows tax").
Actually, check out the latest Gartner group study on the total cost of ownership between Macs and PC's. They found that Macs are about 36% cheaper to run than PC's. And as for the user tax, check out Apples new OSX Server Xserve. No user tax!
Visit Jonesblog and say hello.
> What the Xserve does is incorporate DDR RAM.
/not/ automatic.
...
Well, sort of. The system memory is DDR, for sure, but the connection between the "northbridge" (or whatever Apple calls it) and the processors is still shared 1.06Gb/s PC/133. Ouch! So the Xserve isn't going to be any less memory starved than the current Quicksilver Powermacs, although anything DMA would be better.
And the jury's still out on whether the G4 is really memory starved, anyway. Sure, getting a fatter pipe to the Altivec would be huge, but only for code that can reasonably be rewritten to take advantage of it, and, while certainly there might be big wins in there, it's
Anyway. I'm waiting for MWNY, to see if Motorola can roll out a DDR capable G4. Or, in my wildest dreams, a DDR TiBook
Best,
'jfb
To spur "enterprise Linux," Big Bang, the distributed two-phase commit.
Bill: Steve, we're really concerned that Apple continue development of these programs on Windows.
Steve: Bill, we're really concerned that Microsoft continue development of Office on the Mac.
Bill: (feeling a little bit more of that monopolistic power slip away) DOH!
--Rick
--Rick "If it isn't broken, take it apart and find out why."
...it's not going to be as fast. In fact, MacOS slows some things down for some reason. Mandrake PPC on the G4 I have at work here does slightly better (15% or so) on distributed.net's client than my PIII-600- and it's got a 500 MHz processor. The MacOS version of the cruncher that uses the Altivec is nearly 3.5 times faster than that.
It's all in what you're using of the CPU.
I am not merely a "consumer" or a "taxpayer". I am a Citizen of the State of Texas
"With both companies held by Apple, who will fill the void in the Windows and Linux?"
/.'ers ready to follow them into the Abyss with C# - stop yer whining, already.
I know i didn't just read that.
The idea that Apple would work to assume control of a imperceptably tiny market like this is hardly something to warrant such a whiny question as the poster made.
With the power of 5 black holes crunched together, MS wields the controls of every computer market from the desktop OS, to fscking two button mice. And with a large number of
Apple creeping into a market even farther that they pretty much already 0wwn j00 (and no one cares), this is not the end of the world..
this is the beginning for a major shift of a microsocopic percentage of people who claim already less than 5% of the market.
And the funny part is that, it will end up saving those users money in the long run.
Waaah.
guns kill people like spoons make Rosie O'Donnell fat.
Actually, no, he's right. Apple found that it was more cost-effective to port the parts of the Mac Toolbox that QT needed, than it would be for them to port QT to use the native Win32 API.
If you're talking about GUI elements, then I'm with you-- for QuickTime Player. But are you talking about the QuickTime library itself? I don't believe there's any Mac-specific code in there, at least that I know of.
Exactly. I'm glad you understood my point: you can't compare anything with a 1.5 GHz G4 for the simple reason that it doesn't exist. So instead they compared the fastest G4 with a mid-range Athlon (with a similar price). In fact, they even crippled the Athlon by taking half its memory.
Who cares if an 8 MHz Motorola 68000 (that I have in my good ol' Atari ST) is faster than an 8 MHz Pentium 4? They don't make 2.5 GHz 68000s.
The G4 runs at 1 GHz. The Pentium 4 will soon be running at 3 GHz (currently at 2.53). Even if it's only half as fast per clock cycle, in the end it's still faster. You cand disqualify it on the basis that "it runs too fast and it's not fair".
It's not a matter of MHz; I don't care if a CPU runs at 1 MHz or at 10 GHz. It's a matter of how fast it gets the job done. And right now, the fastest Athlons and Pentiums get pretty much any job done faster than the fastest G4.
There are plenty of good reasons to buy a Mac. But raw speed is not one of them.
Will the G5 change that? Maybe. Personally, I doubt it, but competition is always a good thing.
RMN
~~~
The non-Apple compositing market is far from empty, even when you consider only sub-$50k products. I can think of at least 4, including our own Digital Fusion.
Why would anyone engrave "Elbereth"?
So cry foul yourself.
The real question is why you think Apple is doing something wrong?
Point by point:
It seems that Apple is positioning itself for a monopoly.
Having a monopoly isn't wrong, it's abusing the monopoly and hurting the market that's wrong.
I say this because they are trying to gain an unfair advantage over their competitors...
What's an unfair advantage? You mean they are trying to gain an advantage over their competitors?
by making it so that if people want to use compositing software then they also have to buy a Mac.
Why is that an advantage over their competitors? If the competitor runs on 5 flavors of hardware and Shake and RAYZ only run on four (Mac OS X, Linux, IRIX, and Windows), you think Apple has the advantage?
To make my point: Apple's track record with iDVD, iMovie, FCP, and DVD Studio is to buy a product, make a consumable version, lower the price of the product, and polish the product until it glows.
So the 'advantage' Apple gains over the competition:
More refined product.
Easier to use product.
Cheaper to own and purchase and use product.
Disadvantages? It only runs on Mac OS X.
If Apple stays true to form, we'll see the price drop. This is bad? We'll see better integration and performance. This is bad? We'll see increased usability. This is bad?
If, through the actions of Apple, the consumers *win*, what is Apple doing wrong?
GPL Deconstructed
Only an amatuer would focus so much on the hardware. PC's would be faster. It matters, but not that much. In the recent article about ILM , what were they upgrading from? SGI O2's to Quadro 2's and those were supposed to be 5 times faster. Can you imagine how much better their work will be now?
This Wiki Feeds You TV and Anime - vidwiki.org
And on code that takes advantage of SSE2 you can see a huge improvement on Pentium 4s over Athlons. And so on, and so on.
But new code doesn't grow on trees. Someone has to write it. And even if there's someone around to write it (which often there isn't, especially in closed-source software), a lot of operations don't benefit the least bit from the Altivec, or SSE2, or 3DNow!, or MMX.
There's no point in speculating about code that doesn't exist and optimisations that cannot be made. Vector operations are useful in very specific situations, they're not a magic powder that you can sprinkle over any program to make it faster.
You can only compare things that actually exist, in their current form.
Intel wrote a Pentium 4 optimisation pack for 3D Studio MAX and it turned out that some of those optimisations also helped Athlon systems. But it did make the Pentium 4 more competitive (actually faster, in certain operations). And that's a healthy attitude: show what you product can do and take on the competition. AMD is free to do the same thing, but they probably won't (at least until the Hammer is released).
But Apple's attitude is simply to kill the other versions so they have no competition (and naturally this gives them even less incentive to improve the Mac version). I can see what's in it for Apple (just as I can understand why Microsoft behaves the way it does), but I definitely can't see what's in it for the end users.
Anyway, as I wrote before, I don't use RayZ, (I'm a Discreet / Eyeon kind of guy) so this doesn't really bother me. I just wish Apple would give Intel and AMD a run for their money in the hardware side of things instead of going "the Microsoft Way". Even if they didn't manage to be faster, at least they'd drive prices down.
> Actually, check out the latest Gartner group study on the total cost of
> ownership between Macs and PC's. They found that Macs are about
> 36% cheaper to run than PC's.
Do you have a link to that?
RMN
~~~
You don't seem to understand what more speed means. I doesn't mean more free time. It means more options.
What's important is not rendering a sequence in half the time. It's having enough time to render two slightly different sequences, and picking the one that works best.
So I have no doubt that with a system that's five times as fast, ILM artists will do a better job. Perhaps not five times better (how do you measure quality anyway?), but definitely better.
Nothing is ever finished until the deadline. But on a slow system, it's not finished because it's not finished. On a fast system, it's not finished because you still have time to improve it.
RMN
~~~
Okay, not all of Hollywood and not just Hollywood but you get the idea.
Apple has the editing and they've just bought the compositing. What's left? Well, there's 3D. There has been talk of Apple buying Alias|Wavefront. What about NewTek to get LightWave?
Don't forget that Pixar has RenderMan. Pixar used to want to get out of the software business. What about selling RenderMan to Apple?
What about spinning all that stuff (editing, compositing, 3D, RenderMan) into it's own company, like Apple did with FileMaker? That way Apple can make sure the Mac is very well supported and the company can continue to support other platforms such as Linux (witness FileMaker on Windows)?
Wild speculation, of course, but that is the funnest kind!
I live ze unknown. I love ze unknown. I am ze unknown.
> Without Shake or Nothing Real, is there software available on
> the Mac that even allows you even consider Apple?
Discreet Combustion and Pinnacle Commotion, for example, run on Macs. So does Shake (there was a Mac version before Apple bought Nothing Real). Do does After Effects (not high-end, but not as low-end as it used to be, either).
Making a Mac version of RayZ or any other program (or optimising an existing version) is a great thing. Competition drives prices down and quality up. But killing existing versions, especially the versions that run on the fastest, cheapest platforms (x86), only makes the software less competitive for existing users and less attractive to (potential) future users.
The rest of your points are valid, which is why I say: show me read-world benchmarks. Stop arguing about Altivec and SSE2 and x86-64 and compare software that exists, running on systems that exist.
For example, Pentium 4s are much faster than Athlons in Lightwave, but slower in 3D Studio MAX. So if you're going to be using Lightwave a lot, a Petnium 4 may be a better deal, even if it is more expensive than an Athlon. On the other hand, if you use MAX a lot, Athlons are probably a better deal.
So maybe in some applications a Mac is faster or cheaper than a PC. But the way to find that out is to compare both versions (and in most cases, they already exist). It's not to kill a product that a lot of people have bought and probably depend on. If you've just invested on 50 x86 Linux render nodes, how would it make you feel to hear that the program you use is going to become Mac-only...? Or even that Linux will continue to be supported, but all new versions are going to be Mac-only?
I've written this before: I can understand that Mac users feel happy because of some new program, or new optimisations (even if they don't use that specific program, some innovations might make their way into software they do use). But I cannot understand why they seem to feel so happy just because the other versions (Windows, Linux, Irix) are getting the axe.
RMN
~~~
unix with a good UI, right?
Liberty uber alles.
I've heard that the Windows port took something over 100 engineer years, and I imagine Linux would take at least as many. That's, VERY rough ballpark, $20M.
But, if Apple opened it up and let the community do the port (I'm sure there's the will), it would cost them substantially less. If they only open up the playback side and leave the authoring closed and on the current platforms it would still benefit Linux and *BSD greatly.
With MacOS X (Darwin) being a flavor of FreeBSD they probably have a large proportion of the work required to make a *BSD/Linux port possible done already.
But, back in the real world, what would be their motive for taking such a move? With MacOS X, they already have a viable (unix like) platform for Quicktime and it works quite well (at least on my B&W G3 and iBook2), so the benefits for Apple of having QT support on Linux/*BSD would be low. Despite this point, I'd still love to have a native QT player on my Slackware desktop.
For very large sequences of operations, scan-line rendering is considerably faster and require many orders of magnitude less memory. There is no way to avoid it. Any compositor not designed to work this way will be limitied to a small number of operations before the image must be written back to disk.
From experiments with Rayz it seemed to be scan-line oriented just like Shake and Nuke.
The people who write compositing software do not right fast code. They rarely make use of the vector processors on the hardware on which they run and the authors of said code frequently don't have the knoweldge to push hardware to its limit. In fact, if the authors of a compositing package wanted to do fast real time compositing they'd turn to Apple first to find out how to do it.
-- SIGFPE