Pixar Eclipses Sun with Linux/Intel

Any word on... by 403Forbidden · 2003-02-09 05:59 · Score: 2, Interesting

How fast they can now render over the old Sun servers?

(imaging a Beowulf cluster of THESE!)

Re:Any word on... by AKnightCowboy · 2003-02-09 07:54 · Score: 1

Kind of ironic that an article about how Pixar is dumping Sun for Linux on Intel boxes has a big old Sun ad in the article. I know, offtopic.
Re:Any word on... by Anonymous Coward · 2003-02-09 09:16 · Score: 5, Informative

Sorry Guys... This article looks to be a bit off base!

-- Not an Official RS response --

I work for Rackspace Managed Hosting. The company the link "Rackspace" references in the C-Net article. This kind of cluster is not consistent with our business. We are most focused on web-centric managed hosting vrs colocation. A rendering cluster is something that, from my experience, we've never done. Also We don't carry Blade servers. C-Mon /. I though you guys did better about checking this kind of thing out! Just because it's on c-net doesn't mean it's accurate. Well kudos to who ever really got this job.

Matthew Montgomery
Rackspace Managed Hosting.
Re:Any word on... by DaveCrim · 2003-02-10 09:11 · Score: 1

I think it should have been RackSaver, not Rackspace. RackSaver makes the BladeRack which can hold either 66 or 88 dual processor blades per rack.

Disclaimer: I do not work for RackSaver, just a satisfied customer.

electric bill.... by sirmalloc · 2003-02-09 06:00 · Score: 3, Interesting

1024 xeon's? jeeze, my electric is $120/mo with one amd and one intel running half the time.

Re:electric bill.... by SN74S181 · 2003-02-09 06:29 · Score: 1

You should look into getting a power factor phase corrector, which might reduce your electric bill. People use them to power machinery that has a large inductive load.

Your AMD box, has a huge inductive load. That big array of fan motors that makes it sound like a jet engine.
Re:electric bill.... by ZorinLynx · 2003-02-09 07:12 · Score: 2, Informative

Residences aren't generally penalized for poor power factor like commercial operations are. Normal residential meters measure true wattage. Even if your power factor is lousy, you'll only get billed for actual watt-hours used.
Re:electric bill.... by Noodleroni · 2003-02-09 07:47 · Score: 1

You're crying about $120 per month? Poor little ol' me pays $350 per month for 2 computers that are on 24/7 and 2 others that are on half the time. :-) Thank goodness Pixar doesn't go through my power company. :-P

BTW, three of the computers are AMD, and the other is a Cyrix (it's from before VIA purchased Cyrix.)

--
Esse quam vederi.
Re:electric bill.... by Polo · 2003-02-09 17:14 · Score: 1

Must be the AMD.

For Around... by viper432 · 2003-02-09 06:02 · Score: 3, Informative

For around $25,000 you too can make Pixar quality movies (+ the cost of those servers). https://renderman.pixar.com/

Re:For Around... by viper432 · 2003-02-09 06:05 · Score: 1

Darn... here is the price list:

https://renderman.pixar.com/products/pricelists/in dex.htm

Didnt directly link to that last time.
Re:For Around... by viper432 · 2003-02-09 06:08 · Score: 2, Informative

One LAST try.

https://renderman.pixar.com/products/pricelists/in dex.htm
Re:For Around... by stu_coates · 2003-02-09 06:09 · Score: 4, Insightful

For around $25,000 you too can make Pixar quality movies

...plus the several million $ for the creative talent!
Re:For Around... by whereiswaldo · 2003-02-09 06:22 · Score: 4, Funny

For around $25,000 you too can make Pixar quality movies

1) Download renderman from https://renderman.pixar.com/
2) Learn how to use it.
3) ???
4) Profit!

Eureka! The missing link is "...plus the several million $ for the creative talent!"!
Re:For Around... by $$$$$exyGal · 2003-02-09 06:23 · Score: 1

and once again, the rocket-high electricity bill. I bet the bill is more than $25,000 a month.
Pixar has port... Did somebody say Portman ?
--naked

--
Very popular slashdot journal for adul
Re:For Around... by Shaddup · 2003-02-09 06:27 · Score: 3, Informative

Odd coincidence:

I loaded up slashdot while waiting for aqsis to compile. Aqsis is a renderman-compliant open source renderer. Kinda like bmrt. I'm testing it out, and hope to use it for a shaders-related assignment for the comp. graph. course I'm taking.

My point is that you don't have to shell out $25,000 if you just want to mess around with renderman.
Re: For Around... by Black+Parrot · 2003-02-09 06:27 · Score: 5, Funny

> > For around $25,000 you too can make Pixar quality movies

> ...plus the several million $ for the creative talent!

And if that's too expensive you can forgo the creative talent and make Star Wars prequels!

--
Sheesh, evil *and* a jerk. -- Jade
Re: For Around... by Alan+Partridge · 2003-02-09 06:52 · Score: 2, Funny

...or matrix sequels.

--
That was classic intercourse!
Re: For Around... by Bastian · 2003-02-09 08:28 · Score: 1

And if that's still too expensive, you can forgo the technical talent, too, and make the sequel to Gladiator!
Re:For Around... by stew77 · 2003-02-09 09:22 · Score: 1

There's also 3delight. It's not a GPL solution but still $0 and available for many platforms.

No, you really don't need PRMan (not Renderman - Renderman is the API, PRMan is Pixar's implementation) to create animation. Pencil and paper will do for quite a long way - if your idea doesn't work as a story board, don't even bother to start the computer.
Re:For Around... by WWWWolf · 2003-02-10 08:29 · Score: 1

Yeah, Aqsis is there, though they say it's slow and stuff. There's also 3Delight free beta out there.
Okay, so there are Renderman renderers. Wish there were also good modellers - I know there are some semi-decent modellers, but, for example, Ayam is pretty hard to grasp for someone who is not familiar with Renderman on file format level. I really wish Blender folks can come up with a cool Renderman integration some day, as of yet the scripts are quite incomplete.
(For what it's worth, if you want to see how far someone with little clue about 3D and no clue about Renderman can get with Blender, see my renderings =)

SCARED by wwwgregcom · 2003-02-09 06:02 · Score: 2, Funny

I am actually scared, to imagine a beowulf cluster of these.

--
What signature defines me as a person?

1024 CPUS? by Pharmboy · 2003-02-09 06:03 · Score: 3, Interesting

My god, I thought they had trouble scaling Linux that far. Seriously. How the hell do you do that when "stock" linux doesnt like 8 CPUs?

--
Tequila: It's not just for breakfast anymore!

Re:1024 CPUS? by fredrikj · 2003-02-09 06:06 · Score: 1

I think they run the CPU:s independently, so that they assign one frame to render to each machine and let them run in parallell...
Re:1024 CPUS? by MerlynEmrys67 · 2003-02-09 06:11 · Score: 5, Informative

Just in case you didn't guess, this is a cluster of Linux servers, not a single server
If you have a task that can be easily partitioned off (oh like each individual frame would be an easy break for this) you can send each task to a different machine allowing you to parellelize the task.
This is a poor mans version of NUMA (Non Uniform Memory Access) created and popularized by Sequent (now a division of IBM) where rather than have a single pool of addressable memory, you have multiple pools of memory, some with very fast access, some with slower.
What I am wondering is what do they do for the cluster cross connect. In large scale cluster environments, this tends to be a significant bottleneck. In large scale clusters you start seeing things like HIPPI, VIA, and soon to be Infiniband... wonder what this is stocked up with

--
I have mod points and I am not afraid to use them
Re:1024 CPUS? by sql*kitten · 2003-02-09 06:20 · Score: 4, Informative

My god, I thought they had trouble scaling Linux that far. Seriously. How the hell do you do that when "stock" linux doesnt like 8 CPUs?

Because it's not a single system image. Rendering movies is easy to parallelize because you don't need to have once scene rendered before you can render the next; all the information you need is in the model file.
Re:1024 CPUS? by devaldez · 2003-02-09 06:22 · Score: 1

Not that I know or nothing...but they are using dual GbE connections. If you work a little with Ethernet drivers, you can make a serious impact on the latency problem in Ethernet.

Ask Yahoo (nee Inktomi)...they use a proprietary protocol on top of GbE to resolve the latency issues.

--
"... but you can love completely without complete understanding." - Norman Maclean, "A River Runs Through It"
Re:1024 CPUS? by dprice · 2003-02-09 06:33 · Score: 4, Informative

My god, I thought they had trouble scaling Linux that far. Seriously. How the hell do you do that when "stock" linux doesnt like 8 CPUs?

I often see this misconception about multiprocessor machines. Some machines have a true tightly coupled multiprocessor architectures with a shared memory space, like big iron machines from SGI, Sun, and HP. These can be used to run a multithreaded process to speed up time-to-solution for a task. The speed-up is subject to the usual Amdahl's Law restrictions. The blade server machines, like Pixar is using, are 'tightly bolted' multiprocessors which share mechanical components and power supplies, but they effectively look like separate computers. Possibly some of the blades have shared multiprocessors, but no more than a 2-4 cpus per blade. Separate instances of the OS run on each blade.

For easy to partition tasks like computer graphic rendering, each frame render task can be run single threaded, and there can be many tasks running at the same time. The time-to-solution for a single rendered frame is not reduced by parallelization, but the overall throughput is increased by multiple tasks.

Nine women cannot make a baby in one month, but nine women can make nine babies in nine months.
Re:1024 CPUS? by Sinical · 2003-02-09 06:53 · Score: 1

It's 1024 individual CPU machines: think large cluster.
Re:1024 CPUS? by anethema · 2003-02-09 08:25 · Score: 1

Yes and like the parent poster said, its lots of computers clustered together, not just one linux os handling 1024 cpus.

--

It's easier to fight for one's principles than to live up to them.
Re:1024 CPUS? by javiercero · 2003-02-09 08:37 · Score: 1

Again, too many buzz words... no f**king idea what you are talking about. NUMA implies single adress space, this system shares no memory at all. Each node runs a separate instance of the OS. NUMA was not created by sequent BTW.
Re:1024 CPUS? by khuber · 2003-02-09 08:43 · Score: 1

like each individual frame would be an easy break for this
No way, I can't believe they'd render one frame per CPU. Those frames are huge. I bet they break up the frames into dozens (hundreds?) of pieces. Why? Because each frame is much larger than cache and you'd want to keep the working set in cache as much as possible. If you rendered 1024 frames in parallel you'd have to wait for the single CPU time of a frame before you see anything. If you subdivide the render work in the other direction, splitting frames, you would see output much sooner.
Pixar's website gives a render time of 6 hours - 96 hours per frame. An old Sun news release gives a frame size of 1536x922 x 48 bit color, that's about 8 megs/frame.
-Kevin
Re:1024 CPUS? by digitalcowboy · 2003-02-09 09:28 · Score: 4, Funny

Nine women cannot make a baby in one month, but nine women can make nine babies in nine months.

Won't Microsoft's soon-to-be-released BabyMaker .NET allow for nine women to make a baby in one month?

I thought I saw a press release about it a while back but can't seem to find a link now.
Re:1024 CPUS? by deunhido · 2003-02-09 09:39 · Score: 2, Funny

Nine women cannot make a baby in one month, but nine women can make nine babies in nine months

Actually, I think it takes nine women and a man, but I may not be up-to-date on the latest technology.
Re:1024 CPUS? by jfisherwa · 2003-02-09 11:44 · Score: 1

Don't forget that massive antialiasing and multiple effects mean each frame will have anywhere from 4-16 passes.
Re:1024 CPUS? by Elwood+P+Dowd · 2003-02-09 11:52 · Score: 1

Huh.

"Ideally one would desire an indefinitely large memory capacity such that any particular ... word would be immediately available ....
We are ... forced to recognize the possibility of constructing a hierarchy of memories, each of which has greater capacity than the preceding, but which is less quickly accessible."

- A. W. Burks, H. H. Goldstine, and J. von Neumann
Preliminary Discussion of the Logical Design of an Electronic Computing Instrument, 1946

As quoted on page 538 of my computer architecture textbook, Computer Organization and Design: The Hardware/Software Interface John L. Hennessy and David A. Patterson, second edition.

I realize that Burks, Goldstine, and von Neumann are talking about cacheing, while you are talking about parallel processing, but I'd suggest that the idea has been floating around for a very long time. You know, since before the first cpu. Also, I like using that quote :)

--

There are no trails. There are no trees out here.
Re:1024 CPUS? by BlueArchon · 2003-02-09 12:27 · Score: 1

Why would they need to break up the frames? Isn't this cluster used for the final rendering? I believe the artists have fast enough workstations to make preview renders. They don't need to render in the final resolution to spot any problems... And usually they are only working at one part of scene at a time, so a partial rendering of the scene would be enough.
Re:1024 CPUS? by oconnorcjo · 2003-02-09 16:23 · Score: 1

My god, I thought they had trouble scaling Linux that far. Seriously. How the hell do you do that when "stock" linux doesnt like 8 CPUs?
Actually Linux runs well on 8 CPU's from what I have read, and has even run on a sun E1000 with 31 CPU's (though it is possible someone has even a higher record than that). Linux can scale well up to about 16 CPU's, but beyond 16, performance is not so good. The kernel developers were trying to optimize the kernel to handle between 4 and 8 CPU's in the 2.4 kernel. The 2.6 kernel is supposed to have improvements that might even make Linux scale well beyond 16 CPU's. One article about Linux scaling from somebody at IBM who researched this is at http://www-1.ibm.com/linux/news/frye.shtml. If you use google, it is easy to find other info that supports what I have written.
The 1024 CPU's is a beowulf cluster and Linux is very good when clustered. Each machine has it's own operating system (linux) and they work together as one machine by having all the machines networked together and each machine is told to do different tasks. Search "beowulf cluster" on google and it will give you everything you want to know.

--
I miss the Karma Whores.
Re:1024 CPUS? by sloanster · 2003-02-09 16:48 · Score: 1

Tee hee, who told you "stock" linux "doesn't like" 8 CPUs?

As I recall, 8-way Linux specweb results are pretty respectable, since late 2000 or so.

Say, didn't SGI just release a million dollar, 64-bit, 64-CPU Linux supercomputer? SGI reports that it's basically a "stock" linux-2.4.19 kernel with a few SGI patches...

LOL! 64-way? So much for the "can't scale" FUD!
Re:1024 CPUS? by MerlynEmrys67 · 2003-02-09 22:33 · Score: 1

Actually, I think it takes nine women and a man, but I may not be up-to-date on the latest technology.
Well there is cloning directly, so no man needed... what cult is doing this now ?

--
I have mod points and I am not afraid to use them
Re:1024 CPUS? by SuiteSisterMary · 2003-02-10 03:55 · Score: 1

To put it in perspective, say you want to render one second of film. That's 24 frames.

If each frame takes two hours to render, then you're looking at 48 hours. But if you've got 24 computers, each one can render one frame, then it takes two hours.

So with a render farm, you plug in as many random computers as you can, you have a central 'switchboard' computer that you submit a rendering request to, it gets farmed out to the next available farm member, and when it's done, it gets dumped into a repository.

--
Vintage computer games and RPG books available. Email me if you're interested.
Re:1024 CPUS? by SuiteSisterMary · 2003-02-10 03:59 · Score: 1

For a 'render farm,' you can't really call it a 'cluster.' You can do it all with shell scripts.

You have one machine that you submit rendering requests to; it acts on a last in, first out basis. As individual machines in the farm become available, they contact the switchboard machine, maybe through NFS, maybe through HTTP/wget, maybe through somethign else, and ask for the next task. It's given the task, that task is marked as 'in progress' on the switch board, and the client computer eventually comes back and says 'done, and here's the file name that I dumped into the repository' or after a while, it's marked as 'not done' again and goes back into the queue, on the assumption that the client computer has crashed.

To add more systems to the farm is simply a matter of plugging them in and installing and croning the 'get a file to render and render it' shell script.

It's not at all like a 'cluster,' actually.

--
Vintage computer games and RPG books available. Email me if you're interested.
Re:1024 CPUS? by Pharmboy · 2003-02-10 10:16 · Score: 1

Actually Linux runs well on 8 CPU's from what I have read, and has even run on a sun E1000 with 31 CPU's (though it is possible someone has even a higher record than that).

Thanks for an intellegent reply. Ironic how hard they are to get on /. nowdays. I did a little research. My experience is with smp only, not clustered. I own 6 dual cpu boxes (all linux, different flavors), from ppro to p3, but i have been pretty lazy lately, and been using rh stock kernels since 2.4 came out, and frankly behind on smp scalablility.

im sure i will get some hell about that from some trolls, but my goal is stable systems, not to prove how 37337 I am. im too old and tired to be KewL. Besides, they are pretty nice kernel builds from my experience.

just out of curiosity, i had THOUGHT there was a problem/bug/feature/limitation with x86 smp systems that limited it to 15 cpus in practice (16 in theory, but there was a reason it couldnt do the 16th cpu). is that still an issue? i dont remember seeing any x86 boxes over 8 cpus anyway.

--
Tequila: It's not just for breakfast anymore!

Power by tiktok · 2003-02-09 06:05 · Score: 5, Funny

With that type of processing power, they should be able to calculate to infinity...and beyond.

Re:Power by Paradise+Pete · 2003-02-09 09:16 · Score: 3, Funny

Oh my.. It's times like this when I wish there was a moderation "+0 Bad Pun."
And your post is -1, Doesn't know what a pun is.

Re:1024 processers by Anonymous Coward · 2003-02-09 06:05 · Score: 1

1024 is even for the rest of us.

Wasn't this already linked from an earlier story? by toxcspdrmn · 2003-02-09 06:06 · Score: 2, Insightful

Didn't I follow the same link from the earlier Rendezvous with Rama story?

--
"E pur si muove!" - attributed to Galileo Galilei, 1564-1642

Raw CPU power by EwokNinja · 2003-02-09 06:06 · Score: 4, Insightful

Perhaps if Sun spent more time getting their processors faster at good cost they wouldn't be losing this kind of ground. Sun took way too long to come out with their UltraSparc III processor and now clustering technology is at the point where it's much cheaper to string together a bunch of commodity PCs than purchase a high end Sun box.

Re:Raw CPU power by ottffssent · 2003-02-09 06:34 · Score: 4, Informative

Sun isn't about raw CPU power. For that we have POWER and x86. Sun is about massive scaling. Sure, 1 POWER4 or P4 or Athlon beats an Ultrasparc. And 8 USIIIs lose out to 8 POWER4s or Xeons or Hammer CPUs. But Intel and AMD drop off at about 8P systems (though ItaniumII can handle larger systems, and Opteron can scale past 8P with a HT bridge), and the POWER architecture scales to hundreds of processors. Sun though can pack a thousand chips in a single system image, with plans to scale to 4096 (IIRC) within the next 2 years.

I'm sure Sun would love to have a high-performance CPU to field against massive clusters being deployed for highly parallelizable tasks such as rendering, but the fact is that's not where their strengths lie. Huge tasks which cannot be efficiently split are what Sun is good at, tasks where superb scalability in terms of both CPU power and memory are an absolute must.

For more, read Ace's Hardware's excellent volume multiprocessor articles:
Part 1
Part 2
Part 3

--
High-speed Road Trip (18.000KPH)
Re:Raw CPU power by SN74S181 · 2003-02-09 06:34 · Score: 1

Sun doesn't have the resources for actual hardware development, now that they've adopted a strategy of revenue growth through suing their competitors.
Re:Raw CPU power by christophersaul · 2003-02-09 06:43 · Score: 1

Pardon? Sun spend the equivalent of about 15% of revenues on R and D (compared to Dell's 1.5%).
Re:Raw CPU power by christophersaul · 2003-02-09 06:45 · Score: 1

But I read on Slashdot that you can do anything on clusters of small Intel machines and Linux!
Re:Raw CPU power by drinkypoo · 2003-02-09 07:12 · Score: 1

The problem is that everyone else is getting better at scaling. IBM's POWER series is a serious threat, and you can bet that POWER5 will be even better than what they have out there now.

The comment you replied to is entirely on track. Scaling is not enough. Other processors concentrated on speed first, and scaling up cheaply (obviously if you put enough work into the backplane you can make anything scale, but there comes a point at which it's no longer cost-effective for 99.44% of the market, like a Cray) just wasn't part of the equation. Now that clustering is getting cheaper and more effective (It used to be that workstation-class machines were puny, they are becoming more comparatively powerful, and have more memory for "large" data sets) it is more important to scale smoothly; THIS is when other companies will put effort into that.

--
"You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
Re:Raw CPU power by Cyno · 2003-02-09 08:10 · Score: 1

Yeah, but Sun has been having serious hardware problems these last few years. Maybe its just me or maybe its just their desktop systems, but their CPU yeilds seem to produce a lot of bad ones that pass their quality tests and fail a year or two later. I am talking about the USIIi, it remains to be seen if the USIII will have these same problems.

I have yet to land a job managing thousands of enterprise servers, so don't bank your purchasing decisions based on my observations, but personally I would probably trust Power 4 (IBM) or Itanium2 (HP) over a Sun for stability and performance these days. I have yet to see any Itanium2 hardware, but I know all of SGI's previous hardware up to the Onyx2/O2000 systems and think their new Altix or whatever kicks ass.

Between that for performance computing and an IBM mainframe what more do we need? Oh, more overpriced Sun Ultra 5s for the desktop that fail after 2 years of low to normal use (cheap harddrives, bad CPUs, bad memory). Needless to say I don't trust Sun anymore.

The Ultra 1/2 systems were great. Most of them are still running. But I bet these nice new blades will fail in a couple years, too. I can't justify the costs, can you?
Re:Raw CPU power by Pseudonym · 2003-02-09 11:06 · Score: 1

E10ks are more about high availability than scaling. With an E10k, you can replace entire boards while the machine is still running. If that's what you need, Sun may be the way to go for your application.

Pixar's old render farm was based on E4500s, which are about scaling.

--
sub f{($f)=@_;print"$f(q{$f});";}f(q{sub f{($f)=@_;print"$f(q{$f});";}f});
Re:Raw CPU power by Big+Jason · 2003-02-09 19:37 · Score: 1

I just DR'ed a CPU/Memory board out of my 15k last weekend, worked quite well. But in general, I'd rather get an outage rather than deal with a failed DR operation.

They are blade servers. by NetJunkie · 2003-02-09 06:07 · Score: 4, Informative

Not 1024 CPUs in one box. Each CPU sits on a "blade" card and acts like a seperate system. It's a bug cluster.

Re:They are blade servers. by Pharmboy · 2003-02-09 06:13 · Score: 1

Not 1024 CPUs in one box. Each CPU sits on a "blade" card and acts like a seperate system. It's a bug cluster

ahhhh. ok. thanks for an intellegent and simple answer. amazing how others will just "god, you dont know everything? wtf is wrong with you!". 'specially here on /.

--
Tequila: It's not just for breakfast anymore!
Re:They are blade servers. by Pharmboy · 2003-02-09 06:38 · Score: 1

lol, ok, i had to reply.

I certainly wasnt trolling and appreciate your reply. I have a few linux servers setup. one for a web server, one for internet router, and a couple for game servers. plus i ssh in to manage other web sites. I am comfortable with linux, but christ, I am not a "beowulf" expert.

more importantly, im the one who decides what OS and applications our company uses. Now, im an older timer at /. so i am used to it, but many geeks end up INSULTING the very people they want to switch to OSS. I don't get paid to know everything about linux, just enough to do basics and hire others.

I honestly believe that SOME (minority) of people who profess they want M$ to die, really want everything to stay so complicated that only THEY understand it. I can expand on this, but I think you get the general idea. They are the ones that actually reply to all that 'increase your penis' spam.

Remember kiddies: slow ass microsoft support is still considered better than asking about linux and getting flamed in an "open" forum. You might not think so, but those of us that make the decisions to purchase DO.

--
Tequila: It's not just for breakfast anymore!
Re:They are blade servers. by Mitchell+Mebane · 2003-02-09 06:53 · Score: 1

Remember kiddies: slow ass microsoft support is still considered better than asking about linux and getting flamed in an "open" forum.

You should try Gentoo. The Gentoo forums are refreshingly nice and helpful.

--

The roots of education are bitter, but the fruit is sweet.
--Aristotle
Re:They are blade servers. by suwain_2 · 2003-02-09 13:39 · Score: 1

It's a bug cluster.
No, that's their old Windows cluster that you're referring to. ;)

--
________________________________________________
suwain_2 :: quality slashdot p
Re:They are blade servers. by oingoboingo · 2003-02-09 18:55 · Score: 1

You should try Gentoo. The Gentoo forums are refreshingly nice and helpful.

Phew, I've just finished a hard 9-to-5 of being a total cunt to newbies on the Debian mailing lists and IRC channels. You mean that now I can unwind by coming home, firing up my Debian kernel 0.99 box, and calling all the newbies on the Gentoo lists "ass-fucking micro$hit loving cum-slurping lesbian mandrake-loving drones"? All fucking right!

Re:1024 processers by Exitthree · 2003-02-09 06:07 · Score: 2, Informative

1024 is 2^10. Computers operate in binary, and 1024 is an "even" number when you consider binary.

Not Xserves? by splattertrousers · 2003-02-09 06:07 · Score: 2, Interesting

I assumed that Apple created their Xserve rack-mounted servers for exactly this purpose: not just for animation studios, but for Pixar in particular (since Steve Jobs runs both companies and does things like selling Pixar DVDs to Apple to give away in promotions, thereby increasing the number of DVDs sold at launch, getting his movies in front of more people, and of course providing more incentive to buy whatever it is he's bundling the DVDs with).

I guess the density of the blade servers is higher than the Mac servers, but it would have been a big boost to the Xserve's credibility if Pixar had chosen to use a ton of them. Perhaps Apple will make a new server (Xblade?) that's more suited to this use. It wouldn't surprise me...

Re:Not Xserves? by I+Am+The+Owl · 2003-02-09 06:15 · Score: 2, Insightful

Are you kidding? Xserves don't have anywhere near the computational horsepower of the Intel hardware. And, at that, they are probably more expensive than the Xeon machines per unit.
I've said it before and I'll say it again: you are not paying for cutting-edge hardware when you buy an Apple. You are paying for easy to use software. Movie production houses which have teams of professional administrators do not need the handholding that OSX Server would provide.

--

--sdem
Re:Not Xserves? by damiam · 2003-02-09 09:00 · Score: 1

Xserves don't have anywhere near the computational horsepower of the Intel hardware.
Rendering is actually a decent strength of the G4's.
you are not paying for cutting-edge hardware when you buy an Apple. You are paying for easy to use software
In that case, I'm sure Steve Jobs, seeing as he's the CEO of both Apple and Pixar, might be willing to give them some heavily discounted Xserves without whatever software it is that makes Macs so Xpensive.
I'm not saying that Pixar shoulda gone with Macs; I'm sure they considered it and found good reasons not to. I'm just saying that Macs would actually have been a decent choice.

--
It's hard to be religious when certain people are never incinerated by bolts of lightning.
Re:Not Xserves? by SlamMan · 2003-02-09 09:12 · Score: 2, Interesting

You hit the nail right on the head: density. Xserves are great and wonderful machines (possibly excluding the god-awerful sound they make), but they just don't compete this blade servers. And I'm assuming they're not supposed to. We all know tthe advanages of using blades for this kind of thing, so it'd be foolhardy to use even 1U devices here.

--
Mod point free since 2001
Re:Not Xserves? by MoneyT · 2003-02-09 10:40 · Score: 1

As someone else pointed out, Pixar using Xserves would have resulted in possible conflict of interest issues since the job was given to a low bidder. Sort of like you can't bid on your own auction.

--
T Money
World Domination with a plastic spoon since 1984
Re:Not Xserves? by Melantha_Bacchae · 2003-02-09 16:25 · Score: 1

I Am The Owl wrote:

> Are you kidding? Xserves don't have anywhere near the
> computational horsepower of the Intel hardware.

Xserves would have equalled or beat Intel hardware had Motorola come through with the G5 when it was supposed to. Motorola didn't, and Apple is rushing around trying to get a next generation CPU as quickly as it can. When Apple does fix this, the Xserve will equal or beat Intel hardware (even if it has to be made of Intel hardware ;). GPN sources indicate that the next gen processor will be out before Nov. 3, 2004, as Godzilla wants 50 G5 Xserves for his 50th birthday. (Mothra got OS X for her 40th birthday in 2001. No one knows if Apple is taking seriously King Ghidora's demand for a three headed, gold tone G4 iMac for his big 40 in 2004.)

> And, at that, they are probably more expensive than the
> Xeon machines per unit.

The big expense for servers is per client licensing (Windows 2000 per client licensing can double the price of a server). The Xserve is sold with unlimited client licensing, so it actually is a fairly inexpensive server, much like a Linux based server.

> I've said it before and I'll say it again: you are not paying
> for cutting-edge hardware when you buy an Apple. You
> are paying for easy to use software.

You are paying for cutting edge hardware (say Firewire 800), but not (at the moment) a cutting edge CPU. You are also paying for cutting edge industrial design, the first Unix ever to make it on the desktop, and, yes, easy to use software that is also innovative enough for Microsoft to copy it every time an Apple programmer so much as sneezes.

> Movie production houses which have teams of
> professional administrators do not need the handholding
> that OSX Server would provide.

True, but Xserve isn't just a server, it can be a workstation. Using a few Xserves for 3D workstations and using a Linux blade server render farm sounds like a pretty good combo. The person creating a model or setting up a scene gets a good, solid, easy to use 3D workstation. The rendering farm can be cheap and not have to waste CPU power or memory on Aqua or X11.

I've seen a CNN business spotlight on Pixar. The OS X logo appeared a lot of times in that little piece, so I imagine they use it for something (and yeah, the other 90% of the logo appearances was Steve being cute and grabbing some free TV advertising).

"What I'm thinking is different from what you are."
Belabera, "Mothra 3" 1998
Re:Not Xserves? by Puu · 2003-02-10 03:35 · Score: 1

"the first Unix ever to make it on the desktop"

What about e.g. *BSD (with KDE/Gnome)? Not a major, commercial OS like MacOS X, but still a desktop Unix.
Re:Not Xserves? by bill_mcgonigle · 2003-02-11 10:07 · Score: 1

I'm not saying that Pixar shoulda gone with Macs; I'm sure they considered it and found good reasons not to. I'm just saying that Macs would actually have been a decent choice.

I would think they would have used Xserves if at all humanly possible, to prevent bad PR for Apple. That they didn't is rather telling - I was considering an Xserve cluster for an upcoming project, but Steve's let me know there's some reason I don't want them that I don't know about. I'll always use a Mac on my desk, but this is a good warning sign not to fill up the data center with 'em.

--
My God, it's Full of Source!
OUTSIDE_IP=$(dig +short my.ip @outsideip.net)

Are you nuts!? by Anonymous Coward · 2003-02-09 06:09 · Score: 4, Funny

"I'd love to see their electric bill "

Dude, they render stuff... would you not prefer to see that...

Electricity.. How about HEAT?!? by beldraen · 2003-02-09 06:09 · Score: 1

I'm curious to how many BTUs that thing is rated for..

--
Bel, the mostly sane.. "Of course I can't see anything! I'm standing on the shoulders of idiots." -- Me

Re:Electricity.. How about HEAT?!? by trmj · 2003-02-09 06:21 · Score: 1

The heat isn't so much an issue in the winter, in fact, they have the employees brew their coffee on these.

But seriously, overheating is a problem with something this big unless it's spread out pretty well for proper airflow. I know first hand.. if I didn't have the grill removed from my car and it wasn't winter, it would have blown up long ago.

--
Work sucked, until it became unemployment, when it became slightly more tolerable. -Tet

Seems a shame really... by TheRaven64 · 2003-02-09 06:10 · Score: 1, Interesting

One of the advantages of *nix is that if you code is well written then you can write once and compile anywhere (except possibly windows), so it's nice to see a company showing that this is possible. It's a shame they chose Xeon's though. The Xeon is the latest in a long line of chips which evolved from the 8086, a fairly nondescript 16-bit chip. It still caries a lot of legacy bagage around with it and so has to run several times faster (and drink several times as much power generating several times as much heat) as a well designed chip in order to go as fast. Intel and AMD are able to sell chips which compete with RISC chips only because they sell enough that they can seriously outspend everyone else on R&D making a dead architecture last just a little longer. It would have been nice if the Alpha had been able to compete with x86 on price when NT4 was released, maybe then we'd have left the x86 ship for good. Whenever I here someone suggest OS X or Solaris on x86 I always wonder what they have against the OS that they want to cripple it by putting it on such an evil architecture.

--
I am TheRaven on Soylent News

Re:Seems a shame really... by lederhosen · 2003-02-09 06:17 · Score: 1

well x86 is fastest.

But why did the use Xeons????

They will not preform better, they will only
be more costly. Make it 2048 with normal p4.
Re:Seems a shame really... by TheRaven64 · 2003-02-09 06:22 · Score: 3, Interesting

I'm actually a little surprised they use general purpose CPUs for this kind of task. I'd have thought that a load of custom DSPs might be faster, and probably cheaper - How about 1 DSP per pixel (About 10 million?). I'm sure that would really zip along, if they could sort out the memory access issues inherent in this kind of application. Ray tracing is perfect for parallel execution, since each pixel really is independent of each other pixel, and each frame is likewise independent.

--
I am TheRaven on Soylent News
Re:Seems a shame really... by SN74S181 · 2003-02-09 06:23 · Score: 1

making a dead architecture last just a little longer.

Don't look now, but the 'dead architecture' isn't x86. It's killed many of the architectures that advocates like you champion. Maybe not for the purest of reasons, but it has. Ever heard of Protected Mode, by the way?

I didn't think so.
Re:Seems a shame really... by Anonymous Coward · 2003-02-09 06:34 · Score: 1, Interesting

But it killed them by mass production and massmarketing, marketing meant to dupe people into thinking that MHz is all that matters. The innovative companies making high quality cpus couldn't keep up with Intel and AMD MHz wars. It is a shame.
Re:Seems a shame really... by TheSunborn · 2003-02-09 06:34 · Score: 1

A good idear until they upgrade from Renderman 11.0 to 11.1 and make some minor changes to the render code :}
Re:Seems a shame really... by TheRaven64 · 2003-02-09 07:00 · Score: 1

Ever heard of Protected Mode, by the way?
Protected mode? Yes, I've heard of it. Although when I started coding for x86 CPUs it was Real Mode all the way. Since then, the architecture has the same instruction set, with a few bells and whistles. The FPU is /still/ stack based <shudder>. It's a really dire programming environment, which wins by bein Microsoft compatible. I'm amazed to see so many Microsoft supporters on /. today.

--
I am TheRaven on Soylent News
Re:Seems a shame really... by be-fan · 2003-02-09 14:42 · Score: 1

What matters, is that a P4 is performance competitive with CPUs costing an order of magnitude more. The P4 is the fastest processor for integer code. It's within 70% the performance of a Power4 in FPU. It has 4.2GB/sec of memory bandwidth. It costs less than $1000. MHz might not matter, but that kind of performance at that kind of price point certainly does!

--
A deep unwavering belief is a sure sign you're missing something...
Re:Seems a shame really... by be-fan · 2003-02-09 14:45 · Score: 1

Actually, in P6-class processors, the stack-based FPU is there only in the interface. Top of stack changes are free. In the Pentium 4, you should be using SSE2 (flat register file) anyway, given you give a damn about the interface at the ASM-level.

--
A deep unwavering belief is a sure sign you're missing something...
Re:Seems a shame really... by donglekey · 2003-02-09 20:24 · Score: 1

Actually in the early begging of CGI and Renderman, this was tried by the CGI research division of Lucasfilm. Commodity hardware eventually caught up with it, hasn't looked back, and that's all she wrote.

--

This Wiki Feeds You TV and Anime - vidwiki.org
Re:Seems a shame really... by dcmeserve · 2003-02-09 21:39 · Score: 2, Interesting

How about 1 DSP per pixel (About 10 million?). I'm sure that would really zip along, if they could sort out the memory access issues inherent in this kind of application.
Aye, there's the rub.
Given that they picked Intel chips over Athlons, and given that they must have carefully compared chip performance on their particular application (i.e. Renderman), that says to me that Renderman is mostly memory-bandwith limited, rather than heavy-math-calculation limited.
I know a little bit about how that program works: it's not a ray-tracer. It does some basic 3d calcs to trace lines from camera to objects, and to subdivide polygons. But to determine the actual color of each pixel, it's mostly a matter of one or a number of texture-map, shadow-map, and other lookups. Each is addressing a small part of a big range of memory -- probably breaks through the cache incessantly. I'd bet that many of the geometric calculations are memory-limited too, due to the absolutely humongous number of objects they put into a given scene (e.g. blue fur).
From what I understand, DSP's are good for 2D image processing -- because the algorithms are fairly standard, and require a lot of signal-processing-like math. For 3D, perhaps the matrix calcs involved in the coordinate transforms could be done in a dsp, but as I described, that's probably not a big enough piece of the puzzle to make it worth it.
To go a little further: dedicated hardware was actually the original goal of Pixar, even before it was split off as its own company. But they noticed that hardware advances were so fast that their designs were getting obsolete fast. (The name "renderman" actually came from a quip by one of their engineers, commenting on how they'd soon be able to design a machine that could fit in a pocket-sized device that you could carry around like a Walkman.) Anyways, they eventually discarded the custom hardware, because their software-only "practice" version was getting quite acceptable performance levels all by itself -- on general-purpose processors.

--
"Orthodoxy is unconsciousness" - Orwell
Re:Seems a shame really... by Puu · 2003-02-10 04:28 · Score: 1

I'm no expert, but I thought PRMan also heavily uses shader programs in addition to (or instead of) static textures, so with complex effects there can be quite a lot of FP math imvolved too.

Thanks for the "Renderman" etymology. :-)
Re:Seems a shame really... by dcmeserve · 2003-02-10 08:08 · Score: 1

I'm no expert, but I thought PRMan also heavily uses shader programs in addition to (or instead of) static textures...
Yes, indeed, programmatic shaders are one of the most important features of the Renerman standard. But I'm betting that, at Pixar, 90% of the time they use texmaps. It just works out to be a lot easier to do it that way -- provided you have an army of artists to generate all those 2D images! (BTW, technically, renderman always uses programmatic shaders; to do texmaps, you use a simple shader that just goes and looks up the color values to set.)
Plus, since it typically isn't doing raytraycing, it needs to do a lot of "shadow map" lookups, to tell whether an object is being directly lit or not. I don't know all the details of that, but it does involve doing a texture-map-like lookup for every light source.
Thanks for the "Renderman" etymology. :-)
You're welcome! I got it from the Advanced Renderman book. There's a few paragraphs on it in the back.

--
"Orthodoxy is unconsciousness" - Orwell
Re:Seems a shame really... by Puu · 2003-02-13 04:09 · Score: 1

Ah. Thanks for the insight!

I was under the impression ... by SuperDuG · 2003-02-09 06:10 · Score: 3, Interesting

That Sun had tried renderman (or whatever they call it) to run on 32 bit processors and it was a horrible disaster. Something about how it seemed more feasible and cost efficient to use Sun until the days in which the competiting 64 bit processors became cheaper.

I could have sworn that the software couldn't run at all in 64 bit. I'm just wondering if they didn't take a step down when they converted 64-bit optimized code to run on regular high cache 32-bit pentiums.

Great for linux and anyone who has half a brain knows that you can make a very nice system from the Intel Xeon chips and Linux. But Sparcs aren't x86's and they certainly don't run the same. I've been running a server off of a pII 400 mhz Xeon with 2 megs cache on it for nearly 4 years now. It's never failed me yet and I have no intentions of upgrading anytime soon, but then again I'm not rendering anything in 3 deminsions either.

Doesn't dreamworks use this type of technology already?

Damned MPAA members ... we hate you because of your strives for world domination, but then you go and support linux ... bastards we just love to hate you.

Lastly I'm really surprised that Pixar didn't go for a server farm of OS X boxen, just goes to show ya, right tool for the job. Maybe they'll throw darwin on their at least.

--
Ignore the "p2p is theft" trolls, they're just uninformed

Re:I was under the impression ... by donglekey · 2003-02-09 07:15 · Score: 4, Informative

The parent post is somewhat misleading and more than a little spotty, but it got modded up, so I feel I should clarify.

That Sun had tried renderman (or whatever they call it) to run on 32 bit processors and it was a horrible disaster. Something about how it seemed more feasible and cost efficient to use Sun until the days in which the competiting 64 bit processors became cheaper.

Renderman is a standard for going exporting frames to a renderer. Pixar's implementation is called Photorealistic renderman. Sun is not involved in this at all. It has run on x86 procs, as well as Linux for quite a while now. Renderers are relativly easy to port, especially from different Unixes. I am not sure if there are speed advantages to 64 bit computers, or if it is just accuracy and memory like always, which is still a big advantage for a renderer. ( can anyone clarify?) I have a PRman rendered image on my desktop right now on my 450 Mhz PIII. The above quote is pretty much completly false.

Doesn't dreamworks use this type of technology already?

The technology is just running off the shelf software and hardware. Different parts of dreamworks do use Linux heavily.

Damned MPAA members ... we hate you because of your strives for world domination, but then you go and support linux ... bastards we just love to hate you.

This is horribly misinformed. I don't have the energy to go into the whole issue here but suffice to say that this is wildly misplaced frustration. First of all, Pixar is not a member of the MPAA. They have a deal with Disney, which is. That attidude would be fitting and understandable with Disney for various reasons, but making Pixar your enemy is just wrong (except when they sued Larry Gritz personally to hold off competition to Renderman). The same goes for Visual Effects companies. ILM, Imageworks, Digital Domain, PDI, Pixar, Rythm and Hues, Weta, etc. are the best thing that's happening to Linux right now. They are so far removed from the wrongdoings of the MPAA its like me blaming someone for crime when their friends dad is part of the NRA. They are doing only good for Linux, and they are not hyprocrites. They do have deals with studios that are intern part of the MPAA. Not everything is perfect, and these issues are not something that they as companies are, should be, or will be concerned about. They are also starting to contribute to Linux, and I am confident more will come as Linux matures in their pipeline. Building up anger towards Visual Effects companies perpetuates the sterotype of free software advocates being zealots without understanding the whole issue.

--

This Wiki Feeds You TV and Anime - vidwiki.org
Re:I was under the impression ... by SuperDuG · 2003-02-09 07:41 · Score: 1

Ahhh ... the joys of ripping apart posts on slashdot.
The parent post is somewhat misleading and more than a little spotty, but it got modded up, so I feel I should clarify.
Obviously the moderators are too stupid here and I'm a complete doofus for stating my opinion, I stated no "fact" anywhere in my post. Spotty was because I wasn't sure, and neither are you.
Renderman is a standard for going exporting frames to a renderer. Pixar's implementation is called Photorealistic renderman. Sun is not involved in this at all. It has run on x86 procs, as well as Linux for quite a while now. Renderers are relativly easy to port, especially from different Unixes. I am not sure if there are speed advantages to 64 bit computers, or if it is just accuracy and memory like always, which is still a big advantage for a renderer. ( can anyone clarify?) I have a PRman rendered image on my desktop right now on my 450 Mhz PIII. The above quote is pretty much completly false.
I didn't say it didn't work. I said it was a DISASTER. As in the productivity loss was so great that it was not feasible at the time for them to change over. And my comment wasn't false, it was completely true, the costs for programers to re-optmize for hardware that was already moving up to 64 bit was silly and not right for the time. Simple math made that clear, if you can't run it faster and cheaper you don't upgrade what essentially isn't broken.
The technology is just running off the shelf software and hardware. Different parts of dreamworks do use Linux heavily.
The ones that do rendering CGI use linux and irix. And there is a push to move from irix to a complete linux shop, obviously the catering crew doesn't need linux.
This is horribly misinformed. I don't have the energy to go into the whole issue here but suffice to say that this is wildly misplaced frustration. First of all, Pixar is not a member of the MPAA. They have a deal with Disney, which is. That attidude would be fitting and understandable with Disney for various reasons, but making Pixar your enemy is just wrong (except when they sued Larry Gritz personally to hold off competition to Renderman). The same goes for Visual Effects companies. ILM, Imageworks, Digital Domain, PDI, Pixar, Rythm and Hues, Weta, etc. are the best thing that's happening to Linux right now. They are so far removed from the wrongdoings of the MPAA its like me blaming someone for crime when their friends dad is part of the NRA. They are doing only good for Linux, and they are not hyprocrites. They do have deals with studios that are intern part of the MPAA. Not everything is perfect, and these issues are not something that they as companies are, should be, or will be concerned about. They are also starting to contribute to Linux, and I am confident more will come as Linux matures in their pipeline. Building up anger towards Visual Effects companies perpetuates the sterotype of free software advocates being zealots without understanding the whole issue.
Here let me sum it up for ya ... do they make movies ... exactly.
"Don't blame the artist for the managers mistakes" bullshit.

--
Ignore the "p2p is theft" trolls, they're just uninformed
Re:I was under the impression ... by Bastian · 2003-02-09 08:20 · Score: 1

Lastly I'm really surprised that Pixar didn't go for a server farm of OS X boxen, just goes to show ya, right tool for the job. Maybe they'll throw darwin on their at least.

Ok. Darwin makes some sense, but OS X? Why the heck would you want to be running a full GUI plus Aqua plus all the other tings that consume great heaping molten piles of memory that make up OS X on a render farm? On a render farm, you want each machine image running the absolute minimum for kernel and applications to keep everything else free for doing lots and lots and lots of floating point calculations.

For that kind of task, G4's just can't compete with Xeons on price/performance, especially when you count in the difference in cache size.
Re:I was under the impression ... by malducin · 2003-02-09 09:43 · Score: 1

As far as Dreamworks you have to make the distinction between Feature Animation (in Glendale) and PDI (in Silicon Valley). Actually not much of a distinction since both use off the shelf and propietary software. PDI is mostly propietary but htye do use Maya and I believe Shake. Feature Animation has also written quite a bit of propietary software for Linux:

DreamWorks Feature Linux and Animation

The case of Pixar is a bit more subtle. Yes they do make movies but they don't distribute them (Pixar might be looking for a new partner after their last picture on their current deal). Disney distributes them. The big studios like Disney, Sony, Paramount and the like, control the distribution channels and it's them that are members of the MPAA. VFX studios are service providers, much like the food caterers or the ones that provide trained animals. If a company that provides animals, say a film like The Bear, Bethoven, Dr. Doolittle, etc. in essence it's because of them that that movie is possible. Would the animal company be an evil member of the MPAA? No of course they provide the service much like Imageworks, ILM and others. Pixar is kinda borderline, they do make films but are not part of the big mafia of movie studios, they still need to get someone to distribute the movies. Heck you could even see their deal as being screwed, after all Disney only distributes and promotes, but after expenses they split everything right down the middle. Pixar could be doing a ton more money. I see why the argument for Pixar = MPAA = bad could be made, but it's a nebulous argument at the best.
Re:I was under the impression ... by be-fan · 2003-02-09 14:47 · Score: 1

The P4 can handle 64-bit SIMD math just fine.

--
A deep unwavering belief is a sure sign you're missing something...
Re:I was under the impression ... by Pathwalker · 2003-02-09 20:16 · Score: 1

Why the heck would you want to be running a full GUI plus Aqua...

You do realize that booting OSX in text only mode is easy - just edit /etc/ttys and comment out the line that starts:
console "/System/Library/CoreServices/loginwindow.app/ Contents/MacOS/loginwindow"
and uncomment the line before it that says:
#console "/usr/libexec/getty std.9600" vt100 on secure

Or, for one time use, enter >console as the user at the log in prompt and no more GUI overhead...

Re:Now lets see if they can make a good movie by oyenstikker · 2003-02-09 06:10 · Score: 1

Hey, I liked Monsters, Inc.

No, I'm not female. Easy killer.

--
The masses are the crack whores of religion.

If Jobs is CEO of both... by Anonymous Coward · 2003-02-09 06:12 · Score: 1, Interesting

why didn't Pixar go with macs .. i keep hearing from mac fanatics how much better they are .. and how powerful they are .. if thats true .. why didn't pixar buy the xserve for the render farm? ... if its not good for the goose .. how can it be good for the gander

Re:If Jobs is CEO of both... by La+Temperanza · 2003-02-09 06:32 · Score: 1

Maybe they don't think they can get the rendering software they need for OS X or Linux/PPC or whatever. Though with a customer like Pixar, I'd think some of the developers would be delighted to arrange a special build. :/

--

--
est modus in rebus
Re:If Jobs is CEO of both... by damiam · 2003-02-09 09:09 · Score: 2, Informative

They write their own rendering software, and ported it to Linux for this switch. I'm sure they could have done a PPC port instead, if that's what was needed.

--
It's hard to be religious when certain people are never incinerated by bolts of lightning.
Re:If Jobs is CEO of both... by malducin · 2003-02-09 09:19 · Score: 1

Well it's also based on customer demand. I'm sure a lot of studios wanted a Linux port. If there were a large number of customers clamoring for a PPC port I'm sure Pixar would port it right away. But as it is, a great deal of important clients are switching to Linux so that's what was done instead.

Rackspace or Rackable? by bmarklein · 2003-02-09 06:15 · Score: 5, Informative

As far as I know Rackspace is a managed hosting company. Rackable Systems makes servers - Yahoo and Google both use them. Anyone know if the article has it wrong, and Pixar is actually using Rackable machines?

Re:Rackspace or Rackable? by malducin · 2003-02-09 09:26 · Score: 1

That's what I thought too (about a possible typo) but who knows. A few days ago I posted a story about how ILM was using AMD systems on their renderfarm from RackSaver, which also does Xeon systems. At least it shows that RackSaver knows how to work with those types of customers, and besides ILM and Pixar are relatively close.
Re:Rackspace or Rackable? by Absoluttt · 2003-02-09 12:44 · Score: 1

I work for Rackspace, and it isn't us. I don't understand why this wasn't checked, but hey I'm all for free advertising!
Re:Rackspace or Rackable? by sharkey · 2003-02-10 01:30 · Score: 1

Actually, according to the article, they're using neither, but rather Racksaver.

--

--
"Outlook not so good." That magic 8-ball knows everything! I'll ask about Exchange Server next.
Re:Rackspace or Rackable? by bmarklein · 2003-02-11 19:20 · Score: 1

The article originally said Rackspace. The author corrected the story after I emailed him about it.
Re:Rackspace or Rackable? by sharkey · 2003-02-12 01:36 · Score: 1

Ahh.

--

--
"Outlook not so good." That magic 8-ball knows everything! I'll ask about Exchange Server next.

Re:Why not Apple? by TheRaven64 · 2003-02-09 06:17 · Score: 2, Interesting

given the G4's that sit inside the box are easily out-performed by Intel/AMD these days.
G4s still have one of the best vector units in town (far better than MMX/SSE, see Ars Technica for more details), and the kind of stuff Pixar is going to be using them for (ray tracing I assume) is perfectly suited to opperation on an AltiVec unit. I wouldn't be surprised if, with properly optimised code, a G4 couldn't replace 2-3 Xeons in this particular application. (No, I'm not saying that a 1GHz G4 Mac is faster than a 3THz P7 before I get any flames)

--
I am TheRaven on Soylent News

But What About The OS? by Dr.+Wu · 2003-02-09 06:18 · Score: 1

They might not have chosen XServes, but I'm wondering what flavor of open-source OS they are running? It's Linux-based, but there didn't seem to be much said beyond that.

The article also insinuates again that a possible Intel/Apple connection may happen in the near future. Perhaps the servers are going to be running a customized OS, maybe even the mythical Marklar (the Intel-compatible version of OS X that has been rumored for some time).

But still, it would have been nice to see the XServe's get a little more press.

Dr. Wu

Re:But What About The OS? by Znonymous+Coward · 2003-02-09 06:27 · Score: 1

Give it a rest man. I'm a huge Apple fan and even I'm getting sick of the word Marklar.

--
Karma: The shiznight, mostly because I am the Drizzle.
Re:But What About The OS? by marklark · 2003-02-09 09:33 · Score: 1

I'm a huge Apple fan and even I don't know yet what I think of the name "Marklar"...

- marklark :^)
Re:But What About The OS? by Puu · 2003-02-10 03:44 · Score: 1

Maybe he meant future Intel-based Xserves, not the Pixar rendering farm. Although I can't understand why Apple would waste resources on such a thing at this point, with IBM's PowerPC 970 arriving soon. Should be nicely performing and low-power for rack servers.

Electric Bill Calculated... by BigBlockMopar · 2003-02-09 06:19 · Score: 3, Informative

1,024 Intel 2.8GHz Xeon processors... I'd love to see their electric bill ;)

Well, ignoring the power requirements of RAM, bus controllers, network adapters, hard disks which are probably used for boot only...

Intel rates these things for 74.0W thermal dissipation, which is a pretty good measure of the electrical power consumed... since, unless something is badly wrong, your Xeon chip will not dissipate energy as light or sound.

74W x 1,024 = 75,776W continuous.

Assume they're on 24/7. Assume a cost of $0.06 per kWh, including distribution, debt retirement, Ontario's capped electric rates, etc.

There are 30 days in the average month. There are 24 hours in the average day [grin]. Therefore, there are 720 hours per month.

720 hours @ 75,776W = 54,558,720kWh.

Just a little over $3.2 million per month.

I'd imagine it's less than that; their electric rate is probably somewhat less based on their consumption. But consider that the depreciation on that hardware is probably a greater monthly expense than the electricity to power it...

I'm glad Linux is ready for Pixar, because Linux sure ain't ready for the desktop.

--
Fire and Meat. Yummy.

Re:Electric Bill Calculated... by Anonymous Coward · 2003-02-09 06:24 · Score: 5, Insightful

isn't

720 hours @ 75,776W = 54,558,720kWh.

actually 54,558,720Wh (watt-hours, not kilowatt hours), which is 54,558kWh

making it not 3.2million, but only $3200 a month?
Re:Electric Bill Calculated... by BigBlockMopar · 2003-02-09 06:29 · Score: 1

actually 54,558,720Wh (watt-hours, not kilowatt hours), which is 54,558kWh, making it not 3.2million, but only $3200 a month?
Sorry, man, you're right. That's what I get for posting before drinking my morning coffee.

--
Fire and Meat. Yummy.
Re:Electric Bill Calculated... by Anonymous Coward · 2003-02-09 06:32 · Score: 2, Funny

All the same, if I were hit by a $3200 monthly electrical bill I'd keel over just as dead as if it were 3.2million :)
Re:Electric Bill Calculated... by BigBlockMopar · 2003-02-09 06:33 · Score: 1

making it not 3.2million, but only $3200 a month?
...and it just got modded up. Oh boy.

--
Fire and Meat. Yummy.
Re:Electric Bill Calculated... by idealego · 2003-02-09 06:36 · Score: 1, Informative

Apparently some people can't do simple math and they still get modded up by everyone else who apparently can't do simple math either.

Common sense should tell you that running 1000 cpu's isn't going to cost you 3.2M/month as that would be $3200 a cpu a month, duh!

Like you said, it's actually $3274/month which wouldn't be a big deal for a company like Pixar.
Re:Electric Bill Calculated... by mcgroarty · 2003-02-09 06:37 · Score: 1, Redundant

3.2 million for a thousand processors?
Raise your hand if you're paying $3,200 a month for a single processor, please. No? Somebody needs to double-check their estimate. :-)
Re:Electric Bill Calculated... by sql*kitten · 2003-02-09 06:40 · Score: 1

Well, ignoring the power requirements of RAM, bus controllers, network adapters, hard disks which are probably used for boot only...

The cost of powering the computers themselves pales compared to cost of the air conditioning for the data centre. That is the single biggest cost, in a data center in New York it even outweighs rent!
Re:Electric Bill Calculated... by aaarrrgggh · 2003-02-09 06:42 · Score: 2

Also add 50% for cooling and you approach $5k. Not really that bad... not even that big of a farm... should be able to fit it in a little over 1,000 sq. ft.
Re:Electric Bill Calculated... by treat · 2003-02-09 06:55 · Score: 1

Also add 50% for cooling and you approach $5k.
Shouldn't cooling require at least 100% of the power that heating does?
Re:Electric Bill Calculated... by arkanes · 2003-02-09 06:56 · Score: 1

I doubt they run the farm 24/7 at full capacity except when they're actually in final production, too, but who knows. 5k a month is still alot less than they pay for the admins on the boxes.
Re:Electric Bill Calculated... by ZorinLynx · 2003-02-09 07:10 · Score: 4, Informative

No, cooling (AKA air conditioning) just moves heat from one location to another. This takes a lot less energy than the actual heat load being moved.

This is why heat pumps are a lot more efficient than resistance heating. Heat pumps move ambient heat outside (yes, there's ambient heat; even on a cold day) into the building, which requires less energy than producing the heat energy directly.
Re:Electric Bill Calculated... by beefguts · 2003-02-09 07:14 · Score: 1

You probably have to double that electric bill. You would have to consider air conditioning to remove that heat from the room, you can't just open a window. When you stack these new dense blades in a tower, you are basically creating a chimney where the top blades are being cooked by all the lower blades. Not really a minor consideration.
Re:Electric Bill Calculated... by Anonymous Coward · 2003-02-09 07:21 · Score: 1, Funny

You haven't replaced THAT Pentium machine yet?
Re:Electric Bill Calculated... by ImpTech · 2003-02-09 08:17 · Score: 2, Interesting

Of course, in addition to that there'll be the demand charge. Unfortunately, I have no idea what the going rate on that is, but my understanding is that for a high-consumption user, that part of the bill is MUCH higher than the regular KWh part. At least, thats what they told me when I used to read meters.

For those who don't know, its basically a charge based on the peak amount of power you drew over the course of the billing cycle. They take that peak (~76KW in this case), and multiply it by a constant (on the order of 10 or 100, depending on the customer), and multiply that by whatever the rate is (I don't think its the same rate as they multiply KWh by). Anyway, they always told us to be extremely careful reading the demand, because being off by even the lowest order digit could cost the customer thousands extra in some cases.
Re:Electric Bill Calculated... by mentin · 2003-02-09 08:18 · Score: 1

cost of the air conditioning for the data centre. .. is the single biggest cost, in a data center in New York it even outweighs rent!
Wow! In that case maybe it makes sense to switch to mobile processors, like Pentium-M, does not it?
With performance only twice less than desktop versions, it consumes 50 times less electricity. So you will need twice the number of processors, but they will still consume 25 times less power.

--
MSDOS: 20+ years without remote hole in the default install
Re:Electric Bill Calculated... by RadioTV · 2003-02-09 09:27 · Score: 1

I agree with all your points, with the exception of "than the traditional hit the reset button that happens with Windows 3-4 times a day". The Windows 2000 Workstation on my desk will run for weeks without a restart. Normally I only restart to install security fixes. Now this isn't as good as my Linux computers, but it isn't 3-4 times a day.

If I had to restart it that many times a day I would probably fix it and not just complain about it.

Don't complain about FUD and then spread it yourself.

--
I have great faith in fools - self confidence my friends call it. - Edgar Allan Poe
Re:Electric Bill Calculated... by RadioTV · 2003-02-09 09:36 · Score: 1

But if you're dealing with a standalone Linux machine and something has eaten X, you have no alternative but to reboot.

The next time try ctrl+alt+backspace to kill X. Not an ideal solution, but doesn't require a reboot.

--
I have great faith in fools - self confidence my friends call it. - Edgar Allan Poe
Re:Electric Bill Calculated... by BigBlockMopar · 2003-02-09 09:49 · Score: 1

The next time try ctrl+alt+backspace to kill X. Not an ideal solution, but doesn't require a reboot.
True, and indeed I do that. I think my concern is that most users aren't gonna know to "startx" again immediately after that.

This is where the Task Manager built into Windows makes control of crashed/hung processes easier, and it's embarrassing that of all the things for Windows to do better than a Unix derivative, it's that user-level process control should be one of them.

--
Fire and Meat. Yummy.
Re:Electric Bill Calculated... by RadioTV · 2003-02-09 10:14 · Score: 1

I see people say this all the time and I wonder what causes Windows to crash so much for them. I have five 98 SE and Me machines networked with high-end (cutting edge) hardware at home for gamming. We will run some pretty intense games (Homeworld, Delta Force Land Warrior, MechWarrior 4, etc.) for hours (some times 12+ hours) without any problems.

I could understand if you are the type of user that burns a CD while browsing the web, chatting on IRC and editing graphics in Photoshop.

This is not flamebait - I am just trying to understand why so many people have such drastically different results then me.

--
I have great faith in fools - self confidence my friends call it. - Edgar Allan Poe
Re:Electric Bill Calculated... by Glonoinha · 2003-02-09 10:26 · Score: 1

Good catch. I was going to say, I have watch the effect of computers on my electric bill and found that each machine was about $10 if left on 24x7 (a few years ago = slower machine.) I could see the new P4 Xeons sucking down about 3x what my older machine was doing (note that my calc's were done with a 15" monitor, helping buffer the differences between that box and a Xeon with lots of RAM) so $30 per machine isn't unreasonable.

$30 x 1024 = about $30k per month. Dang, still doesn't seem to work out.

Then again the price of electricity could be different here, but a single mid range machine left on 24x7 with the monitor on only a few hours a day is about $10 per month.

--
Glonoinha the MebiByte Slayer
Re:Electric Bill Calculated... by RadioTV · 2003-02-09 10:29 · Score: 1

I use both Linux and Windows on the desktop, but a significant portion of my job is support of 60+ users running mostly Windows 2000 (and some XP Pro - but no 9x). A lot of the time it's not as easy as pulling up task manager and killing the offending app. A lot of times this will leave the system in an unstable state (but restartable) and some times it just plain doesn't work. We also run several applications in DOS boxes or telneted into a UNIX box, so I am reluctant to tell people to just kill off applications that aren't responding. Although I must say that it would really suck to have to tell people "type ps -ef | grep winword.exe and get the process ID then type kill -5 that number". But I guess that if they were running Linux I could ssh in and kill it myself.

--
I have great faith in fools - self confidence my friends call it. - Edgar Allan Poe
Re:Electric Bill Calculated... by damien_kane · 2003-02-09 10:54 · Score: 1

I think my concern is that most users aren't gonna know to "startx" again immediately after that.

If they didn't know how to type 'startx' to get into X in the first place, it means they're running by default in runlevel 4. In this case, hitting ctrl+alt+bkspace kills X and it simply respawns, so they don't have to type 'startx', or simply 'X', as linked by default in most X installations (definitely any stock distro that I have seen).
Re:Electric Bill Calculated... by spongman · 2003-02-09 11:23 · Score: 1

rkill.exe is your friend. it's in the NT resource kit.
Re:Electric Bill Calculated... by adamruck · 2003-02-09 17:16 · Score: 1

first you didn't have your coffee... now you had way to much coffee or something

--
Selling software wont make you money, selling a service will.
Re:Electric Bill Calculated... by WPIDalamar · 2003-02-10 06:53 · Score: 1

Um... no.

You can't move "cold" inside. You can only move heat energy out. The whole conversation of energy thing kicks in.

Now, on a warm day (warmer than you want to keep your server room) it will require more energy to move that heat outside.

It's nearly 100% efficient to turn electricity into heat. It is no where near that efficient to move heat around.
Re:Electric Bill Calculated... by iangoldby · 2003-02-12 10:22 · Score: 1

If ctrl+alt+backspace also kills all your other X applications with unsaved work, you might just as well have rebooted.

The key issue when something freezes is to salvage as much of your unsaved work as possible. How long it takes to do that is rarely more than a secondary consideration. I've used Windows NT4.0 intensively at work for over five years and I have never, ever lost work in one application due to the crashing of another.

(And I do run Linux at home because I think it is better in most ways that are important to me.)

What I want to know is by amigaluvr · 2003-02-09 06:19 · Score: 1

Is renderman open source yet?

Linux is one step, making sure they have a completely open system is another

Re:What I want to know is by Daniel+Phillips · 2003-02-09 07:45 · Score: 2, Informative

Is renderman open source yet?

Renderman is a specification, not a product. There are various open-source efforts to implement the renderman specification, but they all seem to be dormant at the moment. See here.

--
Have you got your LWN subscription yet?
Re:What I want to know is by malducin · 2003-02-09 08:59 · Score: 2, Informative

Actually there is one hosted at Sourceforge that is very active, called Aqsis. There were a couple of other projects like gman that never took off, or were just University projects. Aqsis is making good progress:

Aqsis

There are a few other implementations that also run on Linux like AIR, The aforementioned RenderDotC (which I believe Cinesite used), and 3Delight. Hopefully a product like Liquid (from a guy that worked at Weta), which is a Maya to RIB translator (kinda like MTOR) will also take off which could help in making a more powerful combo.
Re:What I want to know is by captaineo · 2003-02-13 18:20 · Score: 1

Aqsis is under active development. It works pretty well for standard things but is rather slow. Not sure about the other open-source efforts.

``New mathematics'' by monadicIO · 2003-02-09 06:20 · Score: 1

1024 is an "even" number when you consider binary.
Today I managed to learn new mathematics....that the even-ness of a number depends on the base it is expressed in. Hmmmm....perhaps the laws of mathematics change regularly, after all!!

--

The law of excluded middle : Either I'm foo or I'm foobar

Re:``New mathematics'' by fizbin · 2003-02-10 04:15 · Score: 1

1024 is an "even" number when you consider binary.
Today I managed to learn new mathematics....that the even-ness of a number depends on the base it is expressed in. Hmmmm....perhaps the laws of mathematics change regularly, after all!!

Well, yes. "1024" is odd when you consider it in base 5. ("1024" base 5 == "139" base 10) Of course, in binary "1024" isn't a valid representation at all...

Sun?? by SuperDuG · 2003-02-09 06:21 · Score: 2, Informative

This Link makes no mention of Renderman running on anything Sun related, I see IRIX windows XP and RedHat mentioned here. Is this Sparc-64 tree of the RedHat??

I must be lost here, but most of these renderfarms I've seen that use Sun products is for network storage solutions, though they're even losing the marketshare these days. I think what people are starting to realize is that just because you paid a whole lot for it, doesn't mean you got "The Best".

Supercomputers of 5 years ago can be built today with computers being thrown away and setup into a computing cluster. Obviously the good old days of 40 trillion dollar super computers paid for by the goernment aren't the super computers of today.

--
Ignore the "p2p is theft" trolls, they're just uninformed

Re:Sun?? by malducin · 2003-02-09 09:10 · Score: 1

Well they had PRMan ported to Linux several years ago, and it certainly ran on SUN machines, even if that port was never released commercially. Sometimes Pixar takes a long time to release some stuff udes internally, like the notorious Deep Shadow maps, which they presented over two years ago and used in Monsters Inc. but just released it on PRMan 11 (probably after lots of customer complaining).

Anyway you can read this old press release from SUN (1997), specifically saying that SUN machines were used in the renderfarm:

PIXAR AND SUN MICROSYSTSTEMS CREATE MORE POWERFUL RENDERFARM FOR PIXAR ANIMATION STUDIOS

If I remember back then they had a suite of test that gave something like a RenderMark number or something like that and they said SUN got the highest marks. Some of their more recent SUN machines were 8 processor racks which also gave them bang in as little space as possible (the old buildings before they moved to Emmeryville were really tight).
Re:Sun?? by levork · 2003-02-09 11:41 · Score: 1

> even if that port was never released commercially

The Solaris port of PRMan *was* released commercially dating back to at least before the 3.7 release. Due to lack of customer demand, we stopped supporting it in PRMan 10.0.

What would be awesome... by joebp · 2003-02-09 06:21 · Score: 1

... would be Tux featuring as an incidental character in Toy Story 3 :)

Re:What would be awesome... by chopkins1 · 2003-02-09 06:36 · Score: 2, Funny

Tux already had a role in TS2 as Wheezy.

Re:Raw CPU power- Exactly! by That_Dan_Guy · 2003-02-09 06:24 · Score: 5, Interesting

I teach MCSE courses down in Chatsworth, recently we got a lot of Engineers from boeing coming over for Windows XP classes. Why? They're dumping all their Sparc Stations and moving to XP on cheap Intel hardware. Its faster, and 2/3s of the applications they need run it already. The last third they were working on.

The IT people I talked to were surprisingly happy with XP so far. These were all Unix only kind of people actually.

The other thing they were doing were looking into dumping their Crays in favor of LINUX clusters. The comments were along the lines of how much faster and cheaper it was to put together a cluster of a 100 cheap Intel boxes than getting a new Cray. That, and they were all already familiar with the unix style interface. On top of it all, the GUI interface (I think they were running Gnome) was so much nicer than CDE on Solaris.

So Sun it getting it from both sides- Cheap Wintel boxes and Cheap Linux boxes. No wonder they finally relented and released Solaris 9 on Intel.

Likely Rackable! by Crypt0pimP · 2003-02-09 06:25 · Score: 5, Informative

They have half-depth 1U boxes. That's right, two servers in 1U, back to back.

Includes space between the two for cabling and cooling.

They specialize in delivering easy to manage (physically) racks of highly commoditized systems.
(I work with them in a reseller relationship)

Imagine a 71U rack(minus 1U for a switch), with 142 boxes, all dual proc. 248 procs in a rack!

Man, I wish they'd put the right link in there.

--
Striving to achieve a lower state of conciousness

Re:Likely Rackable! by Jeff+DeMaagd · 2003-02-09 09:02 · Score: 1

In this case, I don't think the 1U boxes are being used - as blade servers have computer cards that are smaller than what a 1/2 of a 1U slot.

These blades all fit maybe a dozen or more into a common blade chassis which the blade chassis itself is often rack mountable.
Re:Likely Rackable! by rthille · 2003-02-09 17:56 · Score: 1

I don't think you can fit 142 ports in a 1U switch :-)

--
Awesome furniture, accessories and cabinetry in Santa Rosa, CA: http://humanity-home.com/

Sad about Sun by wazootyman · 2003-02-09 06:27 · Score: 1

I swear, I'm going to be very sad if some day high end servers, etc are all running x86 boxes running Linux. I huge Sun Enterprise server is so much cooler than a bunch of cheap-ass boxes running Linux.

Re:Sad about Sun by Puu · 2003-02-10 05:56 · Score: 1

"Particular application" is the key; Sun gear has certain strenghts. There's still some way to go before a single Linux instance can employ 100+ x86 processors effectively.

If Sun gets (Solaris) to the 4096-way they are (AFAIK) targeting with UltraSPARC IV, they have some more breathing room again.

Me too would hate to see Sun as just another IA32/IA64 system vendor...

This is exactly the kind of thing x86 is good at. by MisterP · 2003-02-09 06:28 · Score: 5, Insightful

A lot of people are going to be saying "just one example of how Sun is dying", but coming from a place that runs several hundred Sun machines (and being a Sun fanboy), I can understand why they made this switch. For shere processing power on-the-cheap, the x86 world has had a lead on Sun and other big UNIX vendors for a few years. Having a decent OS (linux) to run on those machines, makes it even easier to switch.

It's about using the right tool for the job, and now that x86/linux/bsd has matured to a point where it can be used for some professional applications, it only makes sense to see things like this happen.

Sun is going to be around for a long time. As many other people have pointed out, they're just retreating somewhat to more a of niche market, where they are the right tool for the job.

Re:Why not Apple? by TheSunborn · 2003-02-09 06:29 · Score: 1

I don't KNOW it but I really think that pixar use 64 or 80 bit FPU calculations to renderer movies and Altivec can't really handle that. Besides Render man is a BIG application, and adding Altivec support for Renderman would cost more then the 1024 cpu clusster they bought to render movies.

Martin Tilsted

Performance versus Stablility by BrianUofR · 2003-02-09 06:32 · Score: 5, Interesting

This is a big win for Linux, and that is cool, but performance is only half the battle.

The executives at my company are very interested in linux, because of the outrageous leap in processing power per dollar, and the reductions in CPU-based licensing costs for software like Oracle is staggering. The concern, though, is stability.

Sun Fire and Enterprise servers are really expensive, but they stay up all the time. Swapping a failed processor or NIC or memory stick without halting the box is really important on a mission-critical server. Likewise, a well built Sun box never panics, and if it ever does, Sun will insist that their engineers look at the crash dump to figure out what went wrong.

I think Linux has won the performance battle, but what about the stability battle? You need to win both to win the war.

Re:Performance versus Stablility by thx2001r · 2003-02-09 07:40 · Score: 1

Well, sure, but what about a clustered solution of X86 computers? X-86 servers are chock full of hot-swappable components and redundancy.

It seems the nature of clustering such computers makes them a stable solution as well. I suppose until Linux on X86 has MUCH more time under its belt it won't be considered as stable a solution as Sun.

--
-Joe
If we're all god's children, what's so special about Jesus? - Jimmy Carr
Re:Performance versus Stablility by Jeremi · 2003-02-09 07:48 · Score: 1

I think Linux has won the performance battle, but what about the stability battle? You need to win both to
win the war.

I imagine a good solution to this might be the cluster approach, where you replace your Sun megalith with a large number of vanilla x86 boxes connected via a high speed network. This allows you to use commodity parts, but also to swap out failed components without taking down the system... the difference is you now swap out an entire PC at a time instead of just a SIMM. This way you get the reliability of the Sun for the price of N cheap PCs.

--

I don't care if it's 90,000 hectares. That lake was not my doing.
Re:Performance versus Stablility by peril · 2003-02-09 07:50 · Score: 1

No matter what, if you have an active cpu that fails on a sun host, the host panics. (It will blacklist the CPU once it reboots, but there's still an issue of that "reboot").

There isn't a "standard" SMP unix out there that wouldn't die a horrible death if you pull a cpu out that was running a critical section of the kernel when it was pulled. (I mean jese, how could it; let me pull yer brain out while yer thinking...)

Sun does let you stop/remove and add hardware on the fly gracefully, but they don't protect you against hard hardware failures thru any type of fault tolerance in the os/hardware.

Sun did offer a fault/tolrant unix platform that would deal with component failure/etc, (not sure how - checkpoints? voting? logging?), but it was a combined h/w s/w solution that was pretty expensive.

I'm not sure if Sun/Linux are really fighting the battle that you speak of (stability/performance).

The battle IMHO is horizontal vs vertical scalablity, sun has won the vertical scalability, but it's expensive.

Is it cheaper to have a 6-node 4-way SMP linux cluster running OPS or a single 6800 running oracle? Dunno.

My best guess is that with hardware advancing so quickly in the x86 world, the only boxes where sun has a huge advantage is the high end (> 8GB RAM).
Re:Performance versus Stablility by StarTux · 2003-02-09 08:03 · Score: 1

So what are you after? A cluster or a single powerful uber machine?

Take some advice...IBM, Sun, HP et al have enterprise departments ready to give you excellent advice and deals based on what you're doing. Of course this should be a different experience than if you are I went to CompUSA and asked for a computer without knowing what you wanted or needed :).

"Time spent in reconnaisence is time seldom wasted", well worth remembering that British Army proverb...

StarTux
Re:Performance versus Stablility by marcmcn · 2003-02-09 08:24 · Score: 1

When Element K jumped whole heartedly to a J2EE environment on Sun boxes they ran backwards. They purchased a broken LMS from what was Isopia (now Sun) and had to rebuild it to run properly, if at all. Hopefully they can unwind the years of MS SQL and ASP running their sites. Too bad they chose to remove so many human assets to pay for all of this non-working and incomplete hardware/software, as well as withdraw their IPO. Maybe they would still have an active training center? Linux would have done wonders for their internal and external concerns. They would have had one system to port to rather than a myriad of isolated systems.
Re:Performance versus Stablility by Magnus+Reftel · 2003-02-09 09:18 · Score: 1

I think Linux has won the performance battle, but what about the stability battle? You need to win both to win the war.
Apparently not

--
print "Yet another p{erl,ython} hacker\n",
Re:Performance versus Stablility by nightfire-unique · 2003-02-09 13:27 · Score: 1

There isn't a "standard" SMP unix out there that wouldn't die a horrible death if you pull a cpu out that was running a critical section of the kernel ...
Well, it appears that Sun has discontinued it, but for some period of time, they offered the Netra ft-1800. It did just that; it was an 8 CPU machine, with 4 available for use by the OS. Instructions were checkpointed, and a failure of a CPU module did not affect the availability of the software running on it. It was pretty cool... wish they still offered it. :)

--
A government is a body of people notably ungoverned - AC
Re:Performance versus Stablility by Ydna · 2003-02-09 15:14 · Score: 1

I mean jese, how could it; let me pull yer brain out while yer thinking...

Well, as long as I'm not thinking, go ahead and yank away.

--
"The great thing about multitasking is that several things can go wrong at once." -me
Re:Performance versus Stablility by hondo77 · 2003-02-10 05:56 · Score: 1

Stability matters for some things but not others. Nodes in a render farm aren't mission critical; if one box drops off it's not that big of a deal because you have a lot of others to fall back on. In that case, you are looking more at bang-for-the-buck. Heck, you don't even need maintenance contracts on these boxes because if you can't revive a box, it's cheaper to just replace it.

--
I live ze unknown. I love ze unknown. I am ze unknown.
Re:Performance versus Stablility by thx2001r · 2003-02-11 08:18 · Score: 1

When was the last time you pulled a CPU out of a SUN machine and (if it was a single processor system as most clustered x86 machines are (assuming you want to compare fairly)) that node did any more processing (during the time the CPU was out) than did any x86 machine with its CPU out?

--
-Joe
If we're all god's children, what's so special about Jesus? - Jimmy Carr
Re:Performance versus Stablility by thx2001r · 2003-02-11 08:22 · Score: 1

Besides, AC, what kind of stability does it demonstrate that SUN machines need to have hot-swappable CPU's?

Does the machine automatically swap the CPU for you? I assume that the machine needs human interaction, just the same as an x86 component of an x86 cluster would. If SUN creates a robot that spots physical problems (and fixes them without human interaction), such as impending CPU failures, before they occur, I will really agree that suns are vastly more stable.

Unfortunately, they don't have that system in place yet so I don't see how a sun machine is in any way more stable than a cluster of redundant x86 machines?

--
-Joe
If we're all god's children, what's so special about Jesus? - Jimmy Carr

Google by mrm677 · 2003-02-09 06:33 · Score: 5, Informative

I recently attended a talk by Google's chief engineer. They have approximately 15,000 x86 machines running Linux at seven data centers in the United States.

Weird failures occur so often, such as disks returning garbage without the controller informing the OS, that Google does a checksum on _every_ data structure in their user-level software. He also talked about how Linux is good enough for them, but it doesn't perform well with respects to I/O under heavy load. He says they like Linux because they have the source-code and that they minimize excessive I/O loads on their machines. Nobody asked why they don't use FreeBSD but I suspect its because Linux has better hardware support and Google builds their own machines with numerous different components based on the latest technology.

Re:Google by speeding_cat · 2003-02-09 06:52 · Score: 1

Weird failures occur so often, such as disks returning garbage without the controller informing the OS, that Google does a checksum on _every_ data structure in their user-level software. He also talked about how Linux is good enough for them, but it doesn't perform well with respects to I/O under heavy load.
These guys are probably using cheap arse IDE controllers. In any case, x86 boxes are so disposable these days that it is a lot cheaper to replace a box with another one than have a $$$ support contract.
Of course, for this to work flawlessly frequent failures need to be built into the application, meaning that if some box goes down the system is able to stop using it. This is pretty trivial to do for highly distributed applications such as Google search.
Re:Google by foyle · 2003-02-09 06:52 · Score: 2, Informative

Nobody asked why they don't use FreeBSD but I suspect its because Linux has better hardware support and Google builds their own machines with numerous different components based on the latest technology.
I keep seeing people say that Linux has better hardware support than FreeBSD, but it has not been my experience. In the past year, I've had three machines that Redhat 7.3 and 8.0 refuse to work on. Redhat 7.x installers would choke and the 8.0 installer works but leaves you with an unbootable machine when it finishes. Linux just doesn't get along with the Adaptec AIC-789x controller that was built into the motherboard on these machines. FreeBSD, on the other hand, installs and boots fine without any problems.
Re:Google by Geekboy(Wizard) · 2003-02-09 07:41 · Score: 1

It's been asked. I don't have the article avaliable, but Google basicly said "our apps are Linux centric, and moving them to BSD would be difficult".

I don't think he said outright that he would prefer a BSD solution, but it was implied. [/flamebait=off] ;-)
Re:Google by Alomex · 2003-02-09 07:47 · Score: 5, Interesting

Look, the mean-time-to-failure of a hard drive is 15,000 to 20,000 hours. This means that a hard drive stops working at Goole every hour of every day. Truly 24/7.

If you were to look at their dumpster in the back alley, you'd find about 170 hard drives dunked every week.

Wouldn't you cheksum every data transfer under those conditions too?
Re:Google by tealover · 2003-02-09 07:58 · Score: 1

Right. We'll take your word over Google's chief technology officer. Because your extensive experience and professional analysis most surely trumps that of Google's chief technology officer.

--
-- You see, there would be these conclusions that you could jump to
Re:Google by varaani · 2003-02-09 08:55 · Score: 1

mean-time-to-failure of a hard drive is 15,000 to 20,000 hours

No, much more than that. A 120 GB WD Caviar for example is rated for 500 000 hours. While these MTBF figures do not mean that a single drive would last that long, it gives an estimate of the relative failure rate of new drives. When the drive gets older, it fill fail more often, but in the beginning of its life, the failure rate is much lower than your figures would suggest. So Google hardware guys are changing the failed disks only once a day, not once every hour. And when the drives do get old, they're better off changing also working drives for new ones.
Re:Google by Jeff+DeMaagd · 2003-02-09 08:58 · Score: 1

For one, you are talking about _one_ distribution, and differing distributions go about setting their boot-up differently. One distribution not liking your computer does not mean that "Linux doesn't get along with your computer" - all it really means "Red Hat's verion of Linux doesn't get along with your computer".
Re:Google by damiam · 2003-02-09 09:13 · Score: 1

If you read the fucking parent comment, you'll see that Google's CTO never even mentioned FreeBSD, that was the speculation of the poster.

--
It's hard to be religious when certain people are never incinerated by bolts of lightning.
Re:Google by slamb · 2003-02-09 09:14 · Score: 1

I keep seeing people say that Linux has better hardware support than FreeBSD, but it has not been my experience. In the past year, I've had three machines that Redhat 7.3 and 8.0 refuse to work on. Redhat 7.x installers would choke and the 8.0 installer works but leaves you with an unbootable machine when it finishes. Linux just doesn't get along with the Adaptec AIC-789x controller that was built into the motherboard on these machines. FreeBSD, on the other hand, installs and boots fine without any problems.
That's way too specific to draw a general conclusion from. You say three machines, but really the problem is one hardware device. Overall, I don't think anyone can seriously dispute that Linux has much broader hardware support. And I've had no problems with my 2940UW (which I think uses the same chipset), so it's probably something pretty specific about your devices.
Furthermore, you probably could have gotten it to work if you'd really wanted to. Yes, you shouldn't have to do anything more than going through the installer. But typically people are pretty influenced by their bias in this: if they're surprised that it didn't work (as you probably would be for FreeBSD), they'll go the extra mile to find a patch and/or submit a bug report. And then it will work the next time. If they're not surprised (like you for Linux), they won't do that, and it won't be improved. So it's a feedback loop.
Re:Google by Sarcazmo · 2003-02-09 10:09 · Score: 1

These guys are probably using cheap arse IDE controllers.

Well yeah, of course. SCSI would at least double the cost of the machines, if not triple or more. Assuming $500 extra per machine to go SCSI, $500 X 15,000 is 7.5 million dollars. It's a lot cheaper to just pay some lackies $25,000 a year to swap hard disks all day.
Re:Google by captaineo · 2003-02-09 11:17 · Score: 1

The aic78xx driver is kind of a special case - ego conflicts and lapses in communication between Justin Gibbs (the driver author) and the rest of the Linux kernel crew have caused repeated problems with that driver. If nothing works try getting the latest aic78xx driver from Justin's site.

Of course, I understand the silliness of all this - why should Linus or Redhat let internal conflicts cause the quality of released drivers to suffer? Well, such is life in the free software world.
Re:Google by axxackall · 2003-02-09 12:40 · Score: 1

If you are capable to install FreeBSD than you shouldn't choose RH among Linux distros. RH is for end-users. You should choose Gentoo or similar distro. Then you will have to solve all hardware problems in a way as you do in FreeBSD.
But the result will remain the same (bad for BSD): Linux supports more hardware and does it better.
If you need OS for big SMP machine, shall you choose BSD? No, many even BSD people will advise you to stick to Linux for that. If you need a cluster then you'll get same advise.
And I am not going deep into desktop area where BSD's hardware support is just outdated (Firewire, USB, sound, video are just examples).
Sure BSD can work stable on one CPU running nfsd, httpd and sendmail. But that was the achivement 10 years ago - not today.
As for RH installer support of AIC-789x, again, it's not a problem on other distros and even you can do it on RH: just install the system, boot the liveCD with any Linux distro, mount your root, chroot, rebuild and reinstall your kernel.
The hardware support of Linux and the hardware support of RH installer are two different things.

--

Less is more !
Re:Google by sloanster · 2003-02-09 12:50 · Score: 2, Informative

It's a typical noobie mistake to think that redhat = linux. red hat is a popular distro, but it is not the only one, or even the best one for many applications.

I have numerous red hat servers running reliably in production, but have also worked on servers where redhat will not install, or if it manages to install, will not run reliably. But those systems are rock solid today. How? I Installed SuSE Linux, and all is good The folks at SuSE are very very good at getting things right, they pay close attention to detail and are conservative.

If redhat doesn't do the right thing, don't blame linux, try one of the other distros and you'll probably find one that's just right.
Re:Google by skeedlelee · 2003-02-09 13:08 · Score: 1

Well that's just weird. Intuitively, the 15,000-20,000 makes a bit of sense, that's somewhat close to the warranty period these days. And close to the conventional wisdom I've been hearing of the lifetimes for cheap IDE harddrives. However, the drive you mention does quote a 500,000 hr MTBF (although it's qualified with the word 'field') and a Seagate Barracuda I looked up quoted a 600,000 hour MTBF. Anyone know why they would quote a MTBF of >50 years but give a service life of 5 years? It is Mean Time Before Failure right? If they typically fail after >50 years (ignoring the fact that that hasn't been tested exactly), then why suggest that they will only last 5 years or less. I'm guessing here but are they referring to the loss of readable magnetic signal, not the mechanicals in the MTBF?
Re:Google by modicr · 2003-02-09 14:25 · Score: 1

Question: What MTBF means ?

Answer (example):
600000 MTBF means that if you buy 500 Barracudas
and use them for 50 days, one Barracuda will fail
(because 500x50x24 is 600000).

Roman
Re:Google by seanadams.com · 2003-02-09 15:08 · Score: 1

And this is why MTBF is utterly meaningless for hard drives. Imagine if tire tread wear were rated in MTBF. "Well, we drove 5000 sets of firestones aroudn the block one time each, and not one of them went bald!"
Re:Google by Hobbex · 2003-02-09 21:53 · Score: 1

You are right, it is based on bad statistics. The previous poster's calculation only works if you assume that all drive failures have an exponential distribution: that is to say that the probability of failing at any given time (given that it is still working right before) is going to be the same. Of course, motor errors, head failures, bad sectors, etc etc don't have this distribution at all but get more likely as time goes by.
Re:Google by modicr · 2003-02-10 00:19 · Score: 1

BTW, you can get more info at

http://www.bluemax.net/docs/Motherboards/General _I nfo/mtbf.html

or (PDF!)

http://www.zzyzx.com/products/whitepapers/pdf/MT BF _and_availability_primer.pdf

Roman
Re:Google by pmz · 2003-02-10 03:16 · Score: 1

a Seagate Barracuda I looked up quoted a 600,000 hour MTBF.

The Cheetah drive I just bought has a MTBF of 1.2 million hours. That's over 300 years, but the drive has a 5-year warranty. Still, I'm hoping it lasts a long good life.

--
Healthcare article at Kuro5hin
Re:Google by cabbey · 2003-02-10 16:20 · Score: 1

Wouldn't you cheksum every data transfer under those conditions too?

Um, no. At that point I'd use hardware raid and let the hardware figure it out. And I don't mean cheap crap like a promise fake raid, I mean real, industrial strength, RAID-5 controllers like Adaptec makes, or IBM uses in the iSeries or Shark.

Beowulf, basically by jabbo · 2003-02-09 06:35 · Score: 1

You've heard of this phenomenon, yes? Assigning many machines to a task? That's how rendering is done. Machine 1 handles scene 1 or frame 1, machine 2 handles frame 36, blah, blah, blah...

It's not a 1024-CPU box... it's several hundred boxes with one or more CPU's. Hence the term 'render farm', along the lines of 'server farm'...

--
Remember that what's inside of you doesn't matter because nobody can see it.

Re:Beowulf, basically by Ziviyr · 2003-02-09 08:42 · Score: 1

Actually each machine can work on a fraction of a frame. Seen it done with Lightwave and Realsoft 3D.

--

Someone set us up the bomb, so shine we are!

Re:Why not Apple? by stu_coates · 2003-02-09 06:38 · Score: 1

Certainly not a troll... I have 3 macs myself along with several x86 Linux/*BSD boxes and wouldn't give up my macs for anything... it's just the plain truth, the G4 cannot keep up with the latest x86 processors... now if/when the G5 (or whatever IBM's new Power4 derivative is called) comes along, this may change... we can live in hope...

What took them so long? by speeding_cat · 2003-02-09 06:40 · Score: 3, Insightful

Sun uniprocessor performance has been very uncompetitive for quite some time now. I bet they would have switched a long time ago if it was not for the difficulty of porting software from Solaris to Linux. Plus human inertia ...

The worst problem for Sun is once they loose customers to Linux, there is no turning back.
They still hold well in 64-bit area, however, once commodity hardware such as x86-64 gets there, this battle will also be over.

This is the main reason why the company is likely to go down the drain.

The Only Natural Base is e. by BigBlockMopar · 2003-02-09 06:43 · Score: 4, Funny

Today I managed to learn new mathematics....that the even-ness of a number depends on the base it is expressed in. Hmmmm....perhaps the laws of mathematics change regularly, after all!!

The only natural base is e. Man arbitrarily likes whole numbers, nature like real numbers, and e is everywhere.

Therefore, ln (1024) = 6.931471806... which is not an even number.

I suggest therefore that an even number of processors for the render farm is either

e^6 = 403.4287935 or

e^7 = 1,096.633158.

Of course, Intel is wedded to the whole numbers of processors thing, which utterly thwarts mathematical logic and correctness. Their site also runs on IIS, so what other foolishness can you expect? Heathens.

--
Fire and Meat. Yummy.

Re:The Only Natural Base is e. by Anonymous Coward · 2003-02-09 09:27 · Score: 1, Funny

The only natural base is e. Man arbitrarily likes whole numbers, nature like real numbers, and e is everywhere.

This caused a fierce dispute in the early days of the computer when some mathematicans proposed to use base-e logic instead of base-2, with "nit" (natural digit) as the unit of information. The idea was eventually canned and the only remain of the issue is the term "nitpicking", i.e. making a fuss over the smallest possible piece of information.

Re:Why not Apple? by SN74S181 · 2003-02-09 06:44 · Score: 1

You just linked to an article that, judging from the URL, dates from the first quarter of 2000.

Technology doesn't stand still. Hell, maybe I should cite a comparision of the 6809 against the 8085 processor. The 6809 (the Motorola part, standin for the Macintosh whatever) probably beats the 8085 from then, too.

The /. troll HOWTO by aoteoroa · 2003-02-09 06:47 · Score: 1

Well done. You have the skills of a master troll.

If anyone else is looking for a funny/informative/insightful article that outlines tested techniques for slashdot trolling I recommned this article

Re:Why not Apple? by SN74S181 · 2003-02-09 06:47 · Score: 1

Macintosh has always been more about coming up with the best code-name for a product.

Face it: AltiVec sounds cooler than MMX. So Apple as always wins.

I mean, when has 'IBM' (the traditional name to refer to all x86 machines if you're a Mac zealot) been innovative. They still haven't built a computer into a makeup-mirror.

There are 10 types of people in the world... by phreakmonkey · 2003-02-09 06:48 · Score: 5, Funny

... I guess we know which one you are.

Yet another ``math fact'' learnt by monadicIO · 2003-02-09 06:49 · Score: 1

Therefore, ln (1024) = 6.931471806... which is not an even number.
I must really be on a roll today.....

--

The law of excluded middle : Either I'm foo or I'm foobar

What really happens... by Anonymous Coward · 2003-02-09 06:50 · Score: 2, Funny

This cluster is so powerful, when they try to render anything with it, all they get is "42" on the console.

Re:Raw CPU power- Exactly! by christophersaul · 2003-02-09 06:52 · Score: 1

The gui is the last thing on the list when buying a compute cluster! And besides, Gnome (and KDE and others) work fine on Solaris.

Damn You!!! by DeeZee · 2003-02-09 06:54 · Score: 1

Just when good ole' Sun was showing a sparc of life again, here comes naughty dirty Lintel and pulls the rug from underneath them...

Pixar is right on the mark. by alienthoughts · 2003-02-09 06:57 · Score: 4, Informative

Pixar is on the right track. I do ASIC verification, mainly on Sun boxes (fastest USparc IIIs, multi-proccessor, 14GBs memory, etc). Lately, I have been running the exact same jobs on an LSF enabled Linux farm of Intel boxes.
The improvement is 3-4 times speedup ie 8 hour Sun jobs take 2 hours on Intels.
For the price of one dual proccesor Sun workstation, you can get ten Intel boxes running linux.
Not only is the speedup great, I need less licences to run the CAD software (doing multiple regression jobs). Since a license seat per CAD tool can run from 30K to 200K each plus 10% a year maintence fee, the savings are huge.

Changing over to linux was trivial. I like and have used Suns for years and Suns were a major player in this industry. But I firmly believe that this paradigm is going to be a SUN KILLER!

Sun is setting... by JavaJoint · 2003-02-09 07:04 · Score: 2, Interesting

Anyone get the feeling that Sun's "brightest hours" are behind them? As others have mentioned, they're getting hit from the Windows XP side, as well as Linux. If Solaris dwindles as a result of this, and becomes a niche/high-end item, what does this say for HP, SGI, and the rest of Unixen?

I've been thinking in terms what are/will be the Big Three:
Linux, Mac OS X, and that other thing.. uh, Windows XP. I wouldn't bet on traditional Unixen as a growth area, by any means. Won't be long for some companies to become "Unix-free and Windows-free" zones...

Re:Sun is setting... by tgd · 2003-02-09 07:46 · Score: 3, Insightful

I agree. I think whats going to end up happening long term is Windows will take and keep the desktop (I just don't see it happening with Linux, this coming from someone whose used it as his only OS at home for ten years), Linux in the datacenter, and OS-X in the same niche role its in now, with the caveat that I think it'll start pulling away the tiny percentage of people who want to run Unix on their desktop.

Ten years running Linux, and tomorrow morning I'm dropping the bills on one of those spiffy gigahertz 17" iMacs. I want Unix, and I want more functional stability than Linux has ever given me (not OS stability, but stability in terms of what programs I can use to do what, what works with what else, etc... )

Re:Why not Apple? by arkanes · 2003-02-09 07:04 · Score: 1

Most recent benchmarks I've seen show that, even for apps traditionally better on Macs, like Photoshop filters, the new (equivilently priced) Xeon's are able to handily beat the G4s. Sometimes just throwing more horses at is IS the best solution.

Re:Why not Apple? by Edmund+Blackadder · 2003-02-09 07:10 · Score: 1

Well Steve Jobs seems to disagree with you.

For some reason the parent poster reminds of the WW2 japanese troops that kept fighting on in the jungles of southern china even after their emperor surrendered.

-1, Troll by Gothmolly · 2003-02-09 07:12 · Score: 1

'nuff said.

--
I want to delete my account but Slashdot doesn't allow it.

Stability Issues by fateswarm · 2003-02-09 07:15 · Score: 1

People are talking of stability superiority of Solaris over Linux on this matter, when they don't take into account these two simple facts:

a) This is supposed to render stuff, not to be a 24/7 system. I suppose it's booted today, used for a week or 4 days and then shut down again, until they have stuff to render.

b) People forget that Solaris runs on Sun hardware, not Linux hardware. Don't blame Linux for IBM and Intel's faults _please_.

Re:Stability Issues by sloanster · 2003-02-09 13:02 · Score: 1

Someone blurted out:

> People forget that Solaris runs on Sun hardware,
> not Linux hardware.

um..

Linux hardware?

You mean linux hardware as in IBM big iron, S/390 mainframe? true, solaris doesn't run on mainframe class linux hardware.

Oh, wait, you mean linux hardware as in IBM RS6000 RISC boxen right? ah, solaris won't run on those systems either, I'm afraid.

Or, do you mean linux hardware as in Alpha chip based systems? true, solaris won't run on Alpha linux hardware...

Or, do you mean linux hardware as in ARM chipset? hmm, you're right, solaris won't run on ARM linux hardware...

Oh, wait, there is one linux platform - no, two linux platforms, that solaris can also use: x86 and sparc.

Bingo!
Re:Stability Issues by fateswarm · 2003-02-09 17:25 · Score: 1

look smart ass, the post was referring to the other smart asses friends of yours that insist of referring to linux as a i386 OS. it was a mistake to use the phrase "linux hardware" but your ignorant smart ass post had no place.
Re:Stability Issues by sloanster · 2003-02-09 19:11 · Score: 1

I'm poking holes in the lame-o idea that linux is an x86-only OS.

True, it can and does run on lowly garden variety x86 boxes, but IBM recently revealed that it's sold a billion and a half bucks worth of Linux related goods and services - and part of that was on mainframe sales.

Linux isn't some x86-bound OS, it covers a large spectrum, wristwatch to supercomputer, that's all I'm saying.

"Linux sure ain't ready for the desktop" by Idou · 2003-02-09 07:15 · Score: 2, Redundant

You know, I really don't know what the logic is of arguing that. The people who are using Linux on their desktops now know Linux well enough to completely disregard that. I suppose you will scare newbies away until someone gives them a knoppix CD to play with, but MS spends BILLIONS already for that your little rant is insignificant in comparison.

Maybe Linux is more than ready for the desktop, it just isn't ready for your narrow view of what a desktop should be. And it is not that I really care that you are not satsified, but bitching to a bunch of volunteers seems a bit insane, because I don't think they really care that your are not satisfied, either.

Regardless, Linux isn't going away anytime soon (at least not in my lifetime), so why don't you create a project devoted to "making it ready for the desktop according to my definitions" instead of wasting your life away making complaints about the fruits of a VOLUNTEER EFFORT.

Do you complain about the Salvation Army or Goodwill Industries, as well?

--
Sdelat' Ameriku velikoy Snova!

Re:"Linux sure ain't ready for the desktop" by BigBlockMopar · 2003-02-09 07:53 · Score: 3, Insightful

You know, I really don't know what the logic is of arguing that. The people who are using Linux on their desktops now know Linux well enough to completely disregard that. I suppose you will scare newbies away until someone gives them a knoppix CD to play with,
I use Linux on my desktop. It's great. It's beautiful. But it's *still not ready* for the desktop - as in, it's still not ready to compete with Windows - because it's still more comparable with Windows 3.1 than it is with Windows XP.
Maybe Linux is more than ready for the desktop, it just isn't ready for your narrow view of what a desktop should be. And it is not that I really care that you are not satsified, but bitching to a bunch of volunteers seems a bit insane, because I don't think they really care that your are not satisfied, either.
Maybe my viewpoint is narrow. Or maybe I'm as big a power user as you can get without actually *thinking* in C.

Note that I administer my own domain on a server farm of Linux and OpenBSD machines which live in my bedroom.

Primarily, my main desktop is an e-mail drone. If Evolution actually worked (ie. didn't take 8 minutes to exit on my machine), then it would be fine. But without a spellchecker competitive to prevalent software, Linux/KDE or Linux/Gnome doesn't even make a good e-mail drone. The spellchecker is so 1995. I want an underlining spell checker.

Does that give me a narrow viewpoint, because I expect features which I could take for granted among the apps of more estabished operating systems? Apparently.

Your lack of a realistic viewpoint and your immediate dismissal of my page as FUD is symptomatic of what is wrong with the Linux/OS community, and why I'm starting to believe that Linux will never be able to get its shit together enough to be more of a fringe group like Apple users.

Try using Windows 2000 or XP sometime. Look at it from a user's perspective - you know, the sort of idiot who opens e-mail virii and who makes the *bulk* of the computer-using public. From that perspective, Windows is great. It does everything reasonably well, whether you're a newbie or expert. Linux doesn't do that yet, and therefore isn't as good a desktop solution as Windows.

I'm waiting for the day someone can prove me wrong, but until you get some actual real-world experience with what end-users want from their operating systems, you'll still just be a whiny 14-year-old living in Mommy and Daddy's basement.

--
Fire and Meat. Yummy.
Re:"Linux sure ain't ready for the desktop" by nusuth · 2003-02-09 08:29 · Score: 4, Interesting

My friend and my fan, I have to disagree with you on that. Once installed according to requirements of the user, linux is more than enough for any desktop use. But it is not trivial to find which components make the desktop you require, or how can you troubleshoot, upgrade or just add software to linux. These require a bit of expertise.
The most important linux skill is how to use internet for help, not any unix skills. For a newbie, it is a hit or miss affair. He grabs a modern desktop oriented linux, installs it in 30 or less minutes, if all of his hardware are supported and all programs newbie wants are already installed, good news, we have a new linux fan. Chances are, that won't happen.
If something goes wrong, it is best option for linux fans that newbie just forgets the idea, right then. Most probably he now has a functional system but with a non-functional usb mouse, cd burner or a sub-optimal refresh rate. He will want to fix and use the system. It is just the mouse, or the printer, or excel documents. He almost succeeded in this linux thing!
Wrong. He still misses the crucial skill.
He will try to fix it and fail, seek help and fail again, try to skim docs and fail, learn where to seek help and fail, read documents and seek help at the correct place with the correct attitude and if he has some luck, succeed at last. Now we have a brand new whiner instead fo a fan. Worse, he half knows what he is talking about.
Eveyone whines about windows all the time too, but it is not the same thing. We don't want scared potential new users. In case of windows, user already knows how much of that whining is about a real problem, that is not the case with linux.
Solution is aiming higher. Linux has to be considerably easier to use and install than windows because non-techie users just have a lot of experience with windows. Even if the fix isn't optimal, there is always a fix a phonecall to someone you know away. Linux doesn't have nearly the same installed base so is denied the luxury. Linux still requires a crucial skill; it shouldn't.
In some areas (considering desktop) linux already is better than windows and in others, it is not too far behind. But it has to better on all fronts. Till than, linux is not ready. You can argue that had market shares of linux and windows magically flipped, we would be saying windows is not ready. Probably you would be right, too. But market share (or rather, user base) has not magically flipped and that is not irrelevant.
I know, I should have read the grandparent.

--
Gentlemen, you can't fight in here, this is the War Room!
Re:"Linux sure ain't ready for the desktop" by Idou · 2003-02-09 09:12 · Score: 2, Interesting

"you'll still just be a whiny 14-year-old living in Mommy and Daddy's basement."

Or a financial analyst for a leading semiconductor supplier, besides, you are the one who is doing the whining, but I digress . . .

You seem to have established many assumptions about how a desktop should work, one such assumption is desktops, servers, and every other MS product should be a separate system ENTIRELY. Therefore, though Linux is good for the server, it ain't ready for the Desktop. However, I believe the post Internet era changes this completely (ironic that you compare Linux to Windows 3.1 . . .).

The fact of the matter is that, up until open source (to me, synonymous with "the Internet"), all software came as square pegs. This is because square pegs are much easier to produce than customized pegs. Proprietary software, which doesn't utilize the power of the Internet to its fulliest, is limitted to the square peg model. This can work great in niche markets (which is why MS is trying to make niches all over the place), but it depends on controlling all standards within the market, which is increasingly difficult as the Internet progresses and as more and more people learn how to program. And, as we see with adoption of Linux, when you don't have to use a square peg, a lot can be gained.

But, I suppose my biggest argument is based on the fact that the majority of the world does NOT own desktops. The definition of "Desktop" depends on this majority. For these users, who lack the assumptions you have been conditioned with, Linux is already a superior product and is being adopted at a very fast rate. As time passes, thanks to Linux, this majority will gain access to "the Desktop." Of course, since Linux is constantly improving itself, I will never really be able to prove that it was ready when I made this post, but reality tends to be grey, like that. Not so black and white, as some people see it.

However, I am afraid that without a Madonna song playing in the background and a video of someone flying around, I have failed to convince you. Oh, well . . . I can't say I cared much to begin with. Linux is definitely ready for the desktop as far as I am concerned, regardless of what you think.

Disclaimer:
I am not the spokesman of Open Source. Nobody is the the spokesman of Open Source. Using stereotypes is an indication of a simplistic mind struggling to oversimplify a complex world. Using stereotypes for the Open Source community is down right ludicrous. So get a grip and come to terms that people can still share software even if they don't always agree . . .

--
Sdelat' Ameriku velikoy Snova!
Re:"Linux sure ain't ready for the desktop" by MyHair · 2003-02-09 09:27 · Score: 1

Do you complain about the Salvation Army or Goodwill Industries, as well?

Damn Goodwill and Salvation Army! I still see poor people under bridges and graffiti on walls. This is why I don't give to United Way, these agencies just aren't ready for production use.

:-)
Re:"Linux sure ain't ready for the desktop" by StarTux · 2003-02-09 12:08 · Score: 1

"I use Linux on my desktop. It's great. It's beautiful. But it's *still not ready* for the desktop - as in, it's still not ready to compete with Windows - because it's still more comparable with Windows 3.1 than it is with Windows XP."

Erm, no...KDE 3.x is more comparable to XP. Ever looked at Win 3.1? Doesn't sound like it...Quite a few users booted DOS without that Win shell as it took over head.

"Maybe my viewpoint is narrow. Or maybe I'm as big a power user as you can get without actually *thinking* in C."

Define "power user" as I am sure those folks who use Blender, Houdini etc think themselves as power users.

"Note that I administer my own domain on a server farm of Linux and OpenBSD machines which live in my bedroom."

Pretty common nowadays...At least amongst geeks and even see non techs starting to do this now too.

"Your lack of a realistic viewpoint and your immediate dismissal of my page as FUD is symptomatic of what is wrong with the Linux/OS community, and why I'm starting to believe that Linux will never be able to get its shit together enough to be more of a fringe group like Apple users."

A perspective from someone who runs servers in his bedroom vs others who walk, talk and run large systems with real customers?

"Try using Windows 2000 or XP sometime. Look at it from a user's perspective - you know, the sort of idiot who opens e-mail virii and who makes the *bulk* of the computer-using public. From that perspective, Windows is great. It does everything reasonably well, whether you're a newbie or expert. Linux doesn't do that yet, and therefore isn't as good a desktop solution as Windows."

And they hate it as they still get confused...Then they go to a comp store and come home with a Mac and end up confused with that too...Your point is? I find Windows so lacking in configurability and in so many other ways it just annoys me to no end. And before you say anything, yes I have used Mac Os 8.x to Jaguar, Linux extensively (barring 2.5.x kernel) Windows from Dos 5.x Win 3.1 to Win XP (many of those in professional environments).

"I'm waiting for the day someone can prove me wrong, but until you get some actual real-world experience with what end-users want from their operating systems, you'll still just be a whiny 14-year-old living in Mommy and Daddy's basement."

How can you be proved wrong with an opinion piece? You've settled on what you want and want you can do to make it happen. But remember what might be the right tool for the job today could well be not the right tool for the job tomorrow. You should re-read what you wrote, talking about servers in a bedroom and then saying 14 year old whiny kids with computers in a basement makes on wonder about you and your experiencee :).

StarTux
Re:"Linux sure ain't ready for the desktop" by bonch · 2003-02-09 12:31 · Score: 1

However, I am afraid that without a Madonna song playing in the background and a video of someone flying around, I have failed to convince you.

I'm sure that was a crushing blow to his argument.

Oh, well . . . I can't say I cared much to begin with. Linux is definitely ready for the desktop as far as I am concerned, regardless of what you think.

Are you arguing that Linux can function as a desktop or that it is currently ready for mainstream desktops?

If the former, of course. Lots of people use it for their desktops. After years of using different distros and fighting through the hell of setting up Linux, I am now comfortable doing so.

If the latter, come back and talk when X isn't a bitch to set up for non-techies, i.e., most of the population, who have better things to do than make time for playing with an OS as a hobby. Come back when the so-called "desktop environments" aren't slow as hell and feel hacked together on top of X. Or when fonts aren't ugly as hell (and I don't have to add some hack to get them to work). Or when there aren't tons of different windowing libraries all doing their own inconsistencies. And there aren't poorly programmed interfaces for everything (to open a file in Xine, you have to click the "://" button with the tooltip of "MRL Browser"...okay...). Come back when someone can rightly criticize all these things without being told that it's "insane" to criticize the work of volunteers or that they don't care anyway and do their own thing (why bother pretending to care about making a desktop environment for the users then?). I could go on, as I'm sure you well know.

It's been my experience that current Linux offereings are sub-par as a slick, easy-to-use, everyday desktop. If anybody points it out, they get flamed for some reason (or the ever-popular and inneffective-as-ever rebuttal: "If you don't like it, make it yourself.")

When I have time to play around with an OS and do technical things, and spend time in my life looking for library packages to set up some arcane thing that is depended on by something else, and hack together an X config file, and run a bunch of poorly-made apps that either all start with "K" or "G", only to have no professional and useful programs or games to run on my system anyway, I fire up Linux (I've tried Red Hat, Slackware, Mandrake, and LFS over the years).

When I want to actually get things done, I fire up Windows XP. And everything just works. Things are much simpler and I'm more productive instead of spending all my time setting things up to be productive.

I have high standards for my GUI/OS, like the rest of the population does, believe it or not. I wonder when Linux developers will begin to realize that if they want their stuff to be adopted by everyone, they have to strive for those high standards as well instead of settling for what they've got and hiding behind "volunteer work" as an excuse. If that's all you're in it for instead of the goal of actually supplanting the top contenders on the desktop, your work will always remain for hobbyists and niche users, second-best to Windows XP or OS X.

I would love for Linux to topple them all and be the ultimate desktop system, but there's no way in hell it's ready for that right now. Not one bit. It has the potential to, but the software isn't there yet.

Of course, I'm sure there are others who feel differently!
Re:"Linux sure ain't ready for the desktop" by steveha · 2003-02-09 13:55 · Score: 1

Linux/KDE or Linux/Gnome doesn't even make a good e-mail drone. The spellchecker is so 1995. I want an underlining spell checker.

I really don't know where you get this. Evolution has an underlining spell checker, and has for a long time! You need gnome-spell installed on your system, and if you are using Evolution 1.2 you need to go into the settings and enable the spell checker. (On my computer I have a choice of American, British, or Canadian English checking.)

P.S. GNOME is still playing catch-up to become as nice as Windows. But the 2.2 release goes a long way. It's really nice! The desktop is almost there; we just need applications and a good default setup, and even naive users will be happy to use it.

steveha

--
lf(1): it's like ls(1) but sorts filenames by extension, tersely
Re:"Linux sure ain't ready for the desktop" by Rakarra · 2003-02-10 09:14 · Score: 1

Primarily, my main desktop is an e-mail drone. If Evolution actually worked (ie. didn't take 8 minutes to exit on my machine), then it would be fine. But without a spellchecker competitive to prevalent software, Linux/KDE or Linux/Gnome doesn't even make a good e-mail drone. The spellchecker is so 1995. I want an underlining spell checker.
I can understand why you think it should be there.. but why would you -want- an underlining spell checker? In general, they are a bad idea, as they don't improve your writing in any way, and by breaking up and interrupting your train of thought constantly, it makes it much harder to write. I always turn the underline feature off.. it's cool for eye candy, but it wasn't a well-thought out feature. When you're trying to get your thoughts down on paper, you do not want constant interruptions.
Re:"Linux sure ain't ready for the desktop" by BigBlockMopar · 2003-02-11 15:35 · Score: 1

I really don't know where you get this. Evolution has an underlining spell checker, and has for a long time! You need gnome-spell installed on your system, and if you are using Evolution 1.2 you need to go into the settings and enable the spell checker. (On my computer I have a choice of American, British, or Canadian English checking.)
For sure! Evolution is beautiful. It's full-featured and well-planned.

The only problem is with speed. When I installed it on my PIII-500, it took *8 minutes* for it to close the mailboxes and exit when I was finished using it.

*Eight minutes.*

That needs work, and it's far from the only speed issue I encountered with Evolution. And that's why I use KMail - I need to get stuff done with my computers. It's also why I don't feel that Evolution is ready for the big leagues.

If it takes a PIII-500 to power an e-mail client, what would it take to power a CAD workstation with solid modelling?

To be honest, I thought I'd documented the problem on my website?
P.S. GNOME is still playing catch-up to become as nice as Windows. But the 2.2 release goes a long way. It's really nice! The desktop is almost there; we just need applications and a good default setup, and even naive users will be happy to use it.
It's beautiful. So is KDE 3.1. Both are getting there.

Primarily the problems are with un-related applications of poor quality, missing features, etc.

I *am* a Linux evangelist, and to be a good salesman, you really have to believe in your product. I do. Linux will conquer (or konquer) the desktop someday; my worry is that we're currently selling the masses on an unfinished (and not quite ready) product.... which puts us in the same league as Microsoft. (Can anyone say Windows 95A, boys and girls?)

--
Fire and Meat. Yummy.
Re:"Linux sure ain't ready for the desktop" by BigBlockMopar · 2003-02-11 15:47 · Score: 1

I can understand why you think it should be there.. but why would you -want- an underlining spell checker? In general, they are a bad idea, as they don't improve your writing in any way, and by breaking up and interrupting your train of thought constantly, it makes it much harder to write. I always turn the underline feature off.. it's cool for eye candy, but it wasn't a well-thought out feature. When you're trying to get your thoughts down on paper, you do not want constant interruptions.
I find it to be quite the opposite. I stop when I've mistyped a word, or when I think I've misspelled something. With an unerlining spellchecker, I can, without anything more than a glance back at the word to see if it's underlined, decide whether it's okay.

This is quite the opposite to a spellchecker which makes you manually Accept Changes / Ignore every technical word, URL or foreign name you might happen to find in an e-mail which you're sending. Evaluating each word, out of context ("I misspelled Hello? ... No, wait, that's the quoted text, English is Gunther's third language...") is slow and tedious compared to a glance.

For me, it's a good feature. Based on its popularity in predominant desktop operating systems, most end-users seem to agree that it either simplifies life or increases productivity.

Of course, it doesn't need to actually underline; highlighting, reverse video, etc. would all work just as well for me... :)

--
Fire and Meat. Yummy.

Stages of a SUN Microsystems by digitalgimpus · 2003-02-09 07:17 · Score: 5, Funny

1982 - born in Nebula - incorporated with 4 employees

1984 - protostar - NFS is introduced

1995 - main sequence begins - Java Released

1996 - red giant - Using Java technology, NASA engineers develop an interactive application allowing anyone on the Internet to be a "virtual participant" in the space administration's groundbreaking mission to Mars.

SUPERNOVA - Sun battles MS over Java and Windows

Blackhole - TODAY!

References:
http://www.sun.com/aboutsun/coinfo/history.html

Re:Stages of a SUN Microsystems by Dr.+Spork · 2003-02-09 10:32 · Score: 1

Funny, except Microsoft remind me much more of a black hole. They try to assimilate everything, and once you get near them, they don't let you get away. Oh, and remember: time itself slows down as you approach the event horizon of a black hole. Is it any wonder, then, that your computer slows down once it's got Microsoft?
No, MS is the real black hole. They suck, you know it, but you won't escape.
Re:Stages of a SUN Microsystems by dcmeserve · 2003-02-09 21:52 · Score: 1

1982 - born in Nebula - incorporated with 4 employees
What? They're finding porn in nebulas now?
Oh, wait... darn

--
"Orthodoxy is unconsciousness" - Orwell

Re:This is exactly the kind of thing x86 is good a by Foresto · 2003-02-09 07:18 · Score: 2, Interesting

and now that x86/linux/bsd has matured to a point where it can be used for some professional applications

For some reason, I just can't seem to resist nit picking here. BSD has been mature enough for professional use for quite a long time now.

In fact, I seem to remember a time (pre-Solaris) when Sun systems ran a form of BSD.

its called Binary Math by nurb432 · 2003-02-09 07:21 · Score: 1

Get a book on binary.. then you will understand why its 1024 and NOT 1000...

Kids these days.. just dont understand the fundamentals.. and this is a good example of this.

Throw a Z80 at one of these new 'graduates' and tell them to build a comptuer.. they would freak.. 'what, no AGP port?'... ' whats a register'...

--
---- Booth was a patriot ----

Re:Raw CPU power or bandwidth to compute ratio by Anonymous Coward · 2003-02-09 07:23 · Score: 1, Informative

Sun isn't about raw CPU power.

Agreed. Nor are traditional business apps. Most traditional business apps (Sun's bread and butter market) require a higher bandwidth to compute ratio than can be accomplished by regular intel gear.

In these cases, just throwing a higher clock speed at the CPU does not really do a whole lot. The backplane, layers of cache and interconnects need to be faster.

Sun knows this, as do the people that buy higher end servers.

Movie rendering is different, and therefore a Linux cluster may do just fine.

A little suprised actually by afidel · 2003-02-09 07:24 · Score: 1

I would think that they would need more ram per node than blade servers would allow. I mean my friends a graphic arts student and even his modeling jobs frequently bump against 4gig limits. I would have thought two or four way Opterons with gobs of ram would be more up their ally.

--
There are 4 boxes to use in the defense of liberty: soap, ballot, jury, ammo. Use in that order. Starting now.

Re:A little suprised actually by Pseudonym · 2003-02-09 11:22 · Score: 1

Pixar's animation system and renderer have very different architectures than your friend's modeller.

Maya, for example, requires that the whole model, in dependency graph form, fits in memory at once. RenderMan does not. Clever occlusion culling in combination with "delayed read archive" geometry mean that only a small amount of the scene is in memory at any given time.

--
sub f{($f)=@_;print"$f(q{$f});";}f(q{sub f{($f)=@_;print"$f(q{$f});";}f});

Which distro are they running? by MoreDruid · 2003-02-09 07:26 · Score: 1

I'm just wondering which distro they are running... Would it be big playa Red Hat or did they choose the less support-backed versions like Debian?
I couldn't find anything on their site mentioning this either.

--
The best weapon of a dictatorship is secrecy, but the best weapon of a democracy should be the weapon of openness.

The fast migration of Linux can be dangerious. by jellomizer · 2003-02-09 07:27 · Score: 1

Migrating from Microsoft to Linux is one thing. But from Sun, AIX, HPUX,...,...,..,BSD to Linux can be a dangerious activity. The strangth of the Traditional *nixs is the fact that their code has been matured and well tested (Speaking in relitive term there are exceptoions that I am sure almost everyone can point out). These systems have been well design to work very dependability and well integrated to work with its hardware. Speed and Price is not always the only features you need to buy hardware. There is hardware realiably, scalibility, power usage, software availability, support, and other factors as well. most Linux distributions) and Windows has a tendency to change a lot and forces people to relarn the product every new version so every couple of years emploies get a new learning curve to over come, While Traditional *nixs tend to be simular from the previous version so they can can upgraded without give the people a new learning curve. Of course that gets rid of some gee-wiz and would you look at that factor but it does save money in the long run.

--
If something is so important that you feel the need to post it on the internet... It probably isn't that important.

Re:The fast migration of Linux can be dangerious. by iggymanz · 2003-02-09 07:55 · Score: 2, Insightful

Funny, I've administered IRIX, HP/UX, Solaris, SunOS over the past 13 years....and after administering (and doing systems & application programming) for Linux for the past 4 years, I haven't found it to be less reliable than the others. In fact, I note that MORE patches for security and reliability are required for the proprietary OS's than Linux (those "monster patch" CD's for Solaris are HUGE, man!). Also, I see many of the huge proprietary vendors of operating systems are now SELLING Linux, as it erodes the market for their closed source OS offerings.

I would predict that Linux will destroy AIX, IRIX and HP/UX.....and Sun's planned weak CPU offerings for the next 3-4 years will make running Solaris very unattractive.
Re:The fast migration of Linux can be dangerious. by Billly+Gates · 2003-02-09 13:40 · Score: 1

" Funny, I've administered IRIX, HP/UX, Solaris, SunOS over the past 13 years....and after administering (and doing systems & application programming) for Linux for the past 4 years, I haven't found it to be less reliable than the other"

Name one server on the most reliable uptime list that runs Linux?

Not intending to start a flame war but Linux has its place and so does unix when it comes to uptime, reliablity, and scalability. However FreeBSD and BSD/OS are the highest and most reliable rated which is interesting. Especially since FreeBSD is all opensource like Linux. I switched to FreeBSD not because of OS instability but I found most distro's to be buggy, and the apps to be beta quality. I never had problems with FreeBSD. Nothing bleeding edge is added and they are not far behind like Debian. Its a perfect ballance. It also is better for a server because of this.

Speaking of sun dieing....Sun will no justt sit by and watch their hardware market get eaten up from the ground up.

Sun just announced a lintel line of inexpensive servers for only 3-4k starting price. I have a friend of a friend who works for Sun who is testing x86 sun workstations. They plan to offer an intel workstation line based on Linux the second or third quarter! Wait untill this summer and fall for the anouncment of the new x86 workstation line as well as more x86 entry level servers. Their own distro is spefically designed to work with their hardware. Expect it to be supperior quality like their unix line because of this. You wont have to worry if your video card is supported or if your ethernet card will do weird things because of a buggy driver. Corporate customers will love something integrated like this. That is something you will not get with a Dell Lintel workstation because of the integration.

Sun also makes money from their software as well. They will not go anywhere. SGI may go under and HP may leave HP-UX dead and focus on Windows.NET hell and maybe a few x86 lintel servers but sun will offer both low end x86 and their standard high availibility mainframes and mini's to corporate customers. If I were sun I would try to make an advertising campaign for pixar, boeing and other former sun customers who are switching to Dell. Solaris is not going anywhere for high end customers even though Linux is eating up on the low end.

--
http://saveie6.com/
Re:The fast migration of Linux can be dangerious. by iggymanz · 2003-02-10 09:49 · Score: 1

I don't see any Solaris boxes on that list either (nor HP/UX, IRIX, or AIX)

Salvation Army ain't ready for the battlefield by wheany · 2003-02-09 07:28 · Score: 3, Funny

Salvation Army sure ain't ready for the battlefield.

Re:Why not Apple? by Anonymous Coward · 2003-02-09 07:34 · Score: 3, Insightful

According to the article: Intel, Sun and AMD submitted bids, Intel won. Apple did not submit a bid. If you don't bid on a contract, don't expect to win it.

Pixar is looking for the most processing power money can buy. Everyone knows except for a few specific cases Apple hardware is slower then offerings containing Intel or AMD processors. What Apple is good for is pretty interfaces and easy of use, both of which are pretty useless in a renderfarm.

Since Jobs is CEO at both Pixar and Apple, we can sure of the fact Apple knew Pixar was shopping for a new server. Jobs being CEO at both would raise conflict of interest charges, if Pixar went with Apple hardware.

Xserver is also targeted at smaller markets, specifically ones that don't have an army of support workers.

Imagine the Slammer power by Sabalon · 2003-02-09 07:44 · Score: 1

if each of these was running and unpatched version of SQL server :)

Re:Why not Apple? by Alan+Partridge · 2003-02-09 07:47 · Score: 3, Insightful

I rather suspect that their Intel blade system is cramming in more than 2 CPUs per rack unit on average. Apple may yet try a 4 CPU per U configuration, but the current Xserve ain't it. FWIW, 2 x 1.25Ghz G4s would put up a FAIRLY good showing against a single P4 2.8... but 1024 Xserves would take up 1024 rack Units - /48 would give over 21 full racks - a lot of space!

--
That was classic intercourse!

I'd rather have... by Pig+Hogger · 2003-02-09 07:50 · Score: 4, Funny

I'd love to see their electric bill ;)

I'd rather have their HEATING BILL...

Parallelism by rugwuk · 2003-02-09 07:57 · Score: 5, Interesting

Its all about the distinction between shared and distributed memory architectures. Different applications benefit from different types of parralelism which the above architectures provide. If to solve the problem independent chunks of code can be run that require no communication at run time then clearly a blade type solution (distrbiuted memory) is viable, but if the calculations are co dependent on each other and require communication of interrim results then the overhead of communication can quickly become the critical path and shared memory parallelism becomes a better solution. It also depends on the level of parralelilsm built into the implementation of the algorithms inside pixars redering program itself.

--
Its one damn thing before another. (Dick Bird 1999)

Re:Parallelism by tolldog · 2003-02-09 21:30 · Score: 1

I have had to say the same thing everytime a render farm story hits slashdot. Its amazing how much people talk about beowulf and ignore other clustering solutions exist.

The bandwidth for shared memory for rendering would be crazy. We were maxing Maya's renderer out with 2gb of ram. Sharing that over a network would kill a system. Of course, having 2 procs on the same box with real shared memory does much better. And is also why SGI's were the render king for a long time with their NUMA implementation and highspeed busses between boxes.

-Tim

--
-I just work here... how am I supposed to know?
Re:Parallelism by rugwuk · 2003-02-10 00:11 · Score: 1

Mmmm. I find slashdot useful for the headlines but often find some of the technical content in the replies a little lacking, it seems too many fokls like the buzzwords of the computing world failing to know what they really mean! Its all about cool and not about useful :-(

--
Its one damn thing before another. (Dick Bird 1999)

Re:1024 processers by Senjutsu · 2003-02-09 08:00 · Score: 1

1024 is 2^10. Computers operate in binary, and 1024 is an "even" number when you consider binary.

Dude, 1024 is an "even" number when you consider that it's not odd.

As Penguin Computing Once Reminded Us... by Lethyos · 2003-02-09 08:07 · Score: 1

"Scientists predict that one day the Sun will fail."

--
Why bother.

Re:1024 processers by wadetemp · 2003-02-09 08:07 · Score: 3, Informative

Why the hell 1024 procesors?? Why not 1000??

1024 nodes makes a perfect 10 dimension hypercube. Hypercubes can have major advantages for speeding communications within sub-cubes, which can speed certain types of parallelized applications. Also with this architecture you can avoid a central switch system.

However, you would have to buy 10 ethernet cards per machine, which would be hard to pull off with blades, and I can't think of a way off the top of my head why a hypercube would help with frame rendering, It might be a data server locality thing... but either way, they have their reasons.

Penguin Computing by fobside · 2003-02-09 08:09 · Score: 1

Is anyone else reminded of this?

Are you kidding? by asv108 · 2003-02-09 08:10 · Score: 1

Why go from being tied from one vendor to another? Buying an Xserve kills all the advantages of using commodity hardware. Not to mention on a price/performance ratio x86 blows xserve out of the water. Xserves are great for shops were you don't have a skilled administration staff, but obviously this is not the case with Pixar.

Re:Penguin Computing WHOOPS by fobside · 2003-02-09 08:12 · Score: 3, Funny

ooops! Here you go: http://tipatat.com/artworks/eclipse.jpg

And they're next movie... by ilctoh · 2003-02-09 08:16 · Score: 1

TuxStory!

--
How many slashes would a slashdot dot, if a slashdot could dot slashes?

Re:Why not Apple? by TheSunborn · 2003-02-09 08:17 · Score: 1

Well acording to Ars tehhnica it can only do single-precision floating-point which is 32 bit and thus useless for any rendering which is going to be used with movies. They require atleast 64bit.

Martin Tilsted

Re:Why not Apple? by Stonent1 · 2003-02-09 08:21 · Score: 1

They may have a good vector unit, but if you just want to dump your code into another architechture's compiler and not optimize, x86 is a pretty good choice. Now, for finely tuned, hand-crafted non portable code, G4's would be very good.

Re:1024 processers by wadetemp · 2003-02-09 08:28 · Score: 2, Informative

Ignore my comment about 10 ethernet cards per machine... you could avoid that and still build a hypercube.

Surprising choice by Thagg · 2003-02-09 08:43 · Score: 4, Insightful

I have heard from several places that Intel's PR flacks have been flogging this story mercilessly, so it's not too surprising to see it show up in Slashdot. Twice.

To get the inaccuracy out of the way -- RenderMan has been running on Linux for several years now, and I would be surprised if Linux wasn't the dominant platform for RenderMan for quite some time, outside of Pixar of course.

I am really surprised, though, that at this point in time they'd go from 64-bit to 32-bit machines, especially as 64-bit PC-like machines are just becoming available. Why not go with Itanium or the new Hammer? Each of Pixar's movies to date have been gloriously more complex and hard-to-render than the last one -- and while I know that they go to fairly extreme lengths to keep the memory footprint down I would think that they'd be bumping up against the 4GB limit already. If not now, then quite soon.

Perhaps this is just a stopgap to get Nemo finished, even 1024 servers is a fairly small cost. Certainly it would be compared to the RenderMan licenses :)

Every RenderMan user except for Pixar has to look to get the maximum rendering power per CPU, as the licenses are $5,000 and up, while the CPUs are far far cheaper than that. I suppose Pixar's figure of merit is rendering power per dollar or rendering power per BTU (for cooling limited situations), or even render power per ft^2. Still, the 32-bit machines are a baffling choice to me.

thad

ps. My company has a render garden (too small to be a render farm) of a dozen or so Athlons.

--
I love Mondays. On a Monday, anything is possible.

Re:Surprising choice by khuber · 2003-02-09 09:03 · Score: 1

I don't see why 64 bitness would be an advantage here at all. This is a problem that you'd want running out of cache, not system RAM. From the other posts it does sound like Renderman sucks a lot of memory, so I'd be interested to know what the runtime environment is like.
The 4GB limit you give for x86 isn't correct. P4s can address up to 64GB of physical memory (36 bit address). UltraSparc IIIcu can only address 16GB.
-Kevin
Re:Surprising choice by captaineo · 2003-02-09 11:12 · Score: 1

I think PRMan has two characteristics that reduce its memory usage below other renderers: first, the texture tile cache means that only a small, fixed amount of memory (well under 100MB) needs to be devoted to textures. Second, unless you are using raytracing PRMan does not need to keep all the finely-tesselated geometry in memory at once. (although AFAIK geometry is the limiting factor on memory in Pixar's scenes - due to PRMan's bucketing algorithm it does need to keep at least bounding boxes for the whole scene in memory).

Last SIGGRAPH a developer from Pixar hinted that they are working on a new architecture that "streams" all geometry through the pipeline without retaining any of it (sort of like modern OpenGL cards, or setting PRMan's bucket size so that the entire image is in one bucket). This kind of renderer would require only a constant (albeit large) amount of memory for the frame/Z buffer, regardless of scene complexity. It makes sense because future scenes will get more and more complex, but frame resolution seems to be holding steady (there are very few situations where you'd really need more than 4K or even 2K pixels, at least for the next few years...)
Re:Surprising choice by Pseudonym · 2003-02-09 11:15 · Score: 1

As has been mentioned previously, 64 bit has no advantage. Yes, they can support less RAM, but they are going from 14 CPUs per box to 2 CPUs per box. Their new farm architecture actually has more RAM per CPU than the old one.

In addition, Itanium performance for CPU-bound applications is bad. No, let's rephrase that: it's shockingly bad. Part of it is that the Itanium's OSes are not yet stable, especially under load. (We've submitted several bug reports for Linux over the course of the last six months.)

Pixar's figure of merit is almost certainly rendering power per dollar of TCO, where the TCO includes network infrastructure, power, space, air conditioning and physical sysadmins.

--
sub f{($f)=@_;print"$f(q{$f});";}f(q{sub f{($f)=@_;print"$f(q{$f});";}f});
Re:Surprising choice by mrm677 · 2003-02-09 16:04 · Score: 2, Informative

In addition, Itanium performance for CPU-bound applications is bad.

Last time I checked the Spec CPU benchmarks, Itanium2 was the leader for floating-point performance. Check them out...they may not be the leader right now but Itanium2 is no slouch.
Re:Surprising choice by Pseudonym · 2003-02-09 17:27 · Score: 1
Interesting. I wouldn't be surprised on that. However:
- Pixar's server farm started phasing in six months ago. Itanium 2 was not an option.
- Contrary to popular belief, most high-end rendering is not floating point bound, but rather is I/O bound and memory bound.
--
sub f{($f)=@_;print"$f(q{$f});";}f(q{sub f{($f)=@_;print"$f(q{$f});";}f});
Re:Surprising choice by tolldog · 2003-02-09 21:44 · Score: 1

I don't think there is much of an advantage of longer data formats for rendering. I could be wrong, I haven't dug into render code in a long time. I know that addressing more memory is good and probably the only real advantage for 64 bit.

I know that most places never has enough space for a render farm. We had to run fiber offsite for half of ours... and the rest of ours was running alll over the studio, a few racks here, a few racks there. Its amazing how much space 500 1u or 2u boxes take up. And with 50 or some of them being midsize tower desktops....

Also, using the hyperthreading option on P4's can cost more money. Because it shows up as 2x the processors, and some vendors license by proc, not by box, you can take up twice the licenses. The little performance boost is not worth that...

-Tim

--
-I just work here... how am I supposed to know?
Re:Surprising choice by khuber · 2003-02-10 00:54 · Score: 1

Thanks for the clarification!
-Kevin
Re:Surprising choice by Puu · 2003-02-10 06:58 · Score: 1

My company has a render garden (too small to be a render farm) of a dozen or so Athlons.

Then I have a render-pot of one Athlon.

Don't cry, dry your eye! by NDPTAL85 · 2003-02-09 08:48 · Score: 1

I understand how you feel. Everyone has something they want to be a "purist" about. But at some point you have to address the reality of the situation and just capitulate to way the world is.

x86 won.

Its not about the most technically elegant or efficient designs. Its about the last man standing after everyone's been thrown out of the ring.

x86 has defeated PPC, Sparc, Alpha, MIPS, PA-RISC, and it will soon defeat Itanium in the form of x86-64. No one is left to fight the good fight. They all went softly into that good night.

Conform. Comply. Collaborate. Resistance is futile. You will be recompiled.

--
Mac OS X and Windows XP working side by side to fight back the night.

Re:Don't cry, dry your eye! by TheRaven64 · 2003-02-09 10:03 · Score: 1

I'm not sure how you can say any architecture 'won'. The race isn't over yet, and robably never will be. In 50 years time, I am fairly sure we won't be using x86 derived chips. I'm equally sure it won't be PPC or Itanium derived. We may well not be using anything even remotely Von Neuman-like. One of the great things about *nix and Java is that it's becomming less important what hardware your using. Java runs everywhere, although it runs best on the Mac because it's built into the OS rather than tacked on (although it's still much slower than native code) and code written on one *nix can often be directly compiled on another. Run-time compilation is a major research area and since a program written for a VM can be profiled at run time it can be optimised on the fly, and actually run faster than native code (this doesn't happen yet, and probably won't for at least 10 years). The idea of a CPU being a static design is quite quaint, I can imagine the following conversation:
'You meam your CPU just stayed the same all the time? So when you were doing floating point operations it still had an integer unit being idle? Why? It's much easier if the integer unit just becomes another floating point unit.'
'I know, but back then nanotechnology was in its infancy and you could only reprogram the hardware on a very limited scale.'
'And did you really just compile programs once?'
'Well, sensible people would compile them with optimisations for their specific computers.'
'But the code changed the same? It didn't record branch outcomes and optimise itself as it ran? Wow. It must have been really slow.'
'It was, but when people started writing code which could be optimised as it ran, it was even slower and people laughed and said it would never take off.'

Beware of declaring a winner too early. Sometimes the next generation of technology can come from a surprising direction. ('You mean you can make vacuum tubes out of sand?!?')

--
I am TheRaven on Soylent News
Re:Don't cry, dry your eye! by SN74S181 · 2003-02-09 10:44 · Score: 1

One of the great things about *nix and Java is that it's becomming less important what hardware your using.

In the case of UNIX, just so long as it has the legacy vestiges of a TTY and a Time Sharing System model, you're correct.
Re:Don't cry, dry your eye! by NDPTAL85 · 2003-02-09 12:01 · Score: 1

Too early? No one in their right mind is thinking about 50 years from now. *I* wasn't talking about 50 years from now. The only reason to bring up such a long timespan is to say that "anything" can happen. I am talking about the past 15 years, now and the immediate (next 5 years) future. For THAT timeframe x86 has won. Seeing as how R&D budgets are dwindling for the alternative CPU's and the ever rising costs of fabrication plants I do not see that changing anytime soon either.

--
Mac OS X and Windows XP working side by side to fight back the night.

The plural of anecdote is not data by nusuth · 2003-02-09 08:53 · Score: 1

You know, your experience is expected. There are some HW that a particular linux distro cannot recognize or configure while FreeBSD can. There might even be some HW that doesn't work under linux at all, while do work with FreeBSD. Considering BSD licence of FreeBSD and GPL of Linux, that all drivers can be ported from FreeBSD to linux easily but not the other direction, those must be very rarely used hardware though.

Oh, I stole the subj. line from someone else.

--

Gentlemen, you can't fight in here, this is the War Room!

s/even/nice round/g by yerricde · 2003-02-09 08:58 · Score: 1

Today I managed to learn new mathematics....that the even-ness of a number depends on the base it is expressed in. Hmmmm....perhaps the laws of mathematics change regularly, after all!!

Minor point of terminology. Grandparent was mistaken and used the term "even" to refer to "nice round". The number 1,024 is a nice round number in base 2 (1,024 == 2^10), just as 1,000 is a nice round number to a base-10 computer (1,000 = 10^3). Whether a number is even or not depends only on whether or not 2 is present in its prime factorization, but nice roundness of a natural number does depend on the base.

My definition of "nice roundness": A "nice round number" in base b is any natural number n = a * b^p with 1. small a and 2. b not a factor of a.

--
Will I retire or break 10K?

Re:This is exactly the kind of thing x86 is good a by NDPTAL85 · 2003-02-09 08:59 · Score: 1

Incorrect. FreeBSD, the most advanced BSD has had piss poor SMP support until the release of 5.0.

--
Mac OS X and Windows XP working side by side to fight back the night.

You should have seen their Phone Bill by BoomerSooner · 2003-02-09 09:09 · Score: 1

It kills their electric bill. (I was an auditor at the phone company).

It's okay for friends to disagree by Idou · 2003-02-09 09:50 · Score: 1

But you are definitely right about a lot of things. I am by no means telling the developers to stop (not that it would matter if I did, cause they wouldn't;).

I believe you are referring to a group of users more savvy about computers. I am referring to the types that have trouble installing software because they download a file and don't know where to look for it to install it. I am afraid such users make up the majority of desktop users I know. And the only way I have ever been able to get them to install anything is by getting them to use Lindows click'n run. Such users do not function well with Windows because they don't know how to defrag, or to delete temp files every once and a while. So they get to a point where they don't use computers much at all, because their windows box has degraded too badly. In such cases, I see Linux being much more self-sustaining.

Anyway, I don't see any reason for us to argue over "what kind of user makes up the majority of users." We both agree that Linux needs to be better, and that is enough to base a community on.

I don't preach to other users to install Linux on their desktops, but I do constantly hand out knoppix cds.

Regardless of where it is today, I think we all have an idea of where it is going;)

--
Sdelat' Ameriku velikoy Snova!

Re:It's okay for friends to disagree by nusuth · 2003-02-09 10:51 · Score: 1

I believe you are referring to a group of users more savvy about computers.
Indeed. Most of my experince is derived from people who can administer their systems properly but not really system-admin types. They want to just *use* their systems instead of looking after them, though they are capable of both.
OTOH I also installed a few linux systems for completely clueless users at work too. Installing and configuring those to their liking was hard and every single problem was blamed on linux in the process, but the systems have been running almost without problems so far and users are happy now.
But they are not relevant right now, they don't decide what OS to use nor what to deploy. Ultimately, the goal is making *them* productive, but in the process the decision makers must be introducing the systems to them, hence we have to make those guys happier with linux first. I have to confess, except for giving free support and some advocacy, I haven't done much for that interim goal.
I don't preach to other users to install Linux on their desktops,
That really is a nice strategy. I felt the need to respond your original because I fear the brand new whiner half knowing what he talks about more than whiner who just read on /. that linux is not ready for desktop. With mozilla, openoffice in place, kde and gnome somewhat matured (kde more so), xft2 and prelinking ready for mass deployement, much better interactivity coming with 2.6 kernel series, I thing the worst we can do is insisting to techie types linux is ready *right now.* Half a year from now, the experience will be *much* better.
but I do constantly hand out knoppix cds.
Knoppix is awesome for someone to try out linux for it does not demand any comitment. Make sure either they have at least 256MB ram or you are there to prevent dangerous combinations such as kde+OOo. Programs demanding both many dynamic libraries and a good chunk of memory kill knoppix performance compared to an installed linux.
Regardless of where it is today, I think we all have an idea of where it is going;)
Linus said 2003 is desktop linux year, who am I to disagree :)

--
Gentlemen, you can't fight in here, this is the War Room!

Re:Why not Apple? by TheRaven64 · 2003-02-09 09:52 · Score: 1

Absolutely right, and I assume that this is the reason they chose x86. I guess the G4 doesn't cut it with price/performance when you take into account the cost of having to optimise your code for the architecture, still I'd have thought they'd want to optimise it for SSE if they were moving to Xeon, or they're wasting a lot of their potential speed.

--
I am TheRaven on Soylent News

The word is 'replace' by BollocksToThis · 2003-02-09 10:10 · Score: 5, Funny

I just eclipsed Windows with Linux on my home system.

I just eclipsed my old toothbrush with a new one.

I just eclipsed the shit in my ass-crack with toilet paper.

Now, don't I sound FUCKING STUPID? Yes, I do.

--
This sig is part of your complete breakfast.

Re:The word is 'replace' by Zak3056 · 2003-02-09 12:30 · Score: 1

just eclipsed Windows with Linux on my home system.

I just eclipsed my old toothbrush with a new one.

I just eclipsed the shit in my ass-crack with toilet paper.

Now, don't I sound FUCKING STUPID? Yes, I do.

Dude... it's a pun. You know, "Sun," "Eclipse," that sort of thing?

There's a word in the english language that will change you life... Decaf.

--
What part of "shall not be infringed" is so hard to understand?

Re:1024 processers by Exitthree · 2003-02-09 10:11 · Score: 1

As clerification to anyone who is confused by my use of the word even, I used it in context of the original post, which you may or may not have read (it's been modded down, so you might not have seen it.) I use it in the context of even number being a number that is nicely rounded, as the original poster refered to 1000 as a nicely rounded, even, number. I also would like to point out that I put in quotes, to suggest that although it isn't the correct definintion of even, it is the definition used in context of the original post. OK? Now get off my back, I know the difference between even and odd numbers.

Re:This is exactly the kind of thing x86 is good a by Nicolas+MONNET · 2003-02-09 10:12 · Score: 1

Well if you don't need a journalling file system, real SMP support nor decent threading for something to be "professional", then I guess *BSD might be "professional", indeed.

Re:Raw CPU power- Exactly! by Dr.+Spork · 2003-02-09 10:14 · Score: 1

Why do former Unix geeks trade in their work computer and get an XP machine? Why, to play games at work, of course!

Re:71U Rack? by Crypt0pimP · 2003-02-09 10:19 · Score: 1

Actually, you are right. I must have still been hazy after waking up.

Thanks for the correction.

--
Striving to achieve a lower state of conciousness

Imagine, this looks like a Beowulf cluster! by sergeaux · 2003-02-09 11:11 · Score: 1

Oh, well...

that's neat and all by Gavitron_zero · 2003-02-09 11:29 · Score: 2, Funny

but when are they going to spend some money and teach their animators how to model a human that doesn't look like a puppet?

Re:that's neat and all by sharkey · 2003-02-10 01:44 · Score: 1
Quick Poll-
Would you rather see:
- Human Actors
- CGI Model that looks like a puppet
- Jar-Jar Binks
--

--
"Outlook not so good." That magic 8-ball knows everything! I'll ask about Exchange Server next.

Re:Irony by ClosedSource · 2003-02-09 12:32 · Score: 2, Insightful

They use Linux instead of Solaris because the platform is cheaper, thus increasing their profits. Their parent company is opposed to the DVD aspects of Linux because they believe it will reduce their profits. There's no irony there, just a consistent focus on making money.

If you look closely you'll find that no major company supports GPL'd software out of principle, they all do it to make more money.

Re:Irony by femto · 2003-02-09 12:53 · Score: 1

Then it looks like I will have to live without hope! :-)

Re:Raw CPU power- Exactly! by Billly+Gates · 2003-02-09 13:04 · Score: 1

" Why do former Unix geeks trade in their work computer and get an XP machine?"

One Word. MS-office. Enough said.

..and no I am sure its the MCSE IT directors making the decsion and not the actual Unix users. This is how ms gains marketshare. They target our bosses with an integrated TCO solution.....then the cost saved is spent on BSA auditing at their own expense.

--
http://saveie6.com/

Pixar, Steve, and OS X on Intel by bfrog · 2003-02-09 13:12 · Score: 1

I like this quote from a Maccentral article: ...Jobs delivered the morning keynote to Intel's annual sales conference in Las Vegas.

Maybe there's hope for OS X on Intel after all. Goodbye Classic, Goodbye Altivec, Hello to some driver rewrites, but what a great leap in price/performance!

I can't think of a better model to begin this transition with than the XServe.

Re:Boeing running XP??? by That_Dan_Guy · 2003-02-09 13:13 · Score: 1

Umm, actually it has more to do with the application that is running on the OS than the OS itself. The applications they want to run to do their CAD work run just as well on XP as Unix. XP boxes cost a fraction of Sparc Stations. Do the math.

As for Airbus, well, its your choice, but when I was taking my Sofwtare Engineering class the instructor explained that the most error free code on the planet had one error per 10,000 lines of code. Airbus was NOT the most error free code on the planet. It had millions of lines of code. And it has a tendency to overcorrect pilot corrections of the program's errors. Good luck landing if it screws up and the Pilot hasn't the time to turn it off before correcting it manually!

Re:Raw CPU power- Exactly! by That_Dan_Guy · 2003-02-09 13:21 · Score: 1

The driving factor was the APPLICATIONS. They ran way faster on Wintel than Sparc, and they were available. The MS IT team was somewhat unprepared and stingy with giving out information on how to integrate the new WinXP machines into the existing Windows Server infrastructure. The (former) Unix Engineers wanted to do RIS and the like. But they were having to put in their own seperate Win2k Servers and figure out how to do the Domain structure for AD.

So it seemed that the MS-IT team didn't push the Unix people to do Windows. In fact, whenever I go to MS-sponsered seminars there is usually one speaker who blasts the way Boeing does their IT work (they write their own management scripts and treat it like Unix instead of Windows or some such MS-sponsered FUD)

Why not go with Itanium or the new Hammer? by Bruce+Perens · 2003-02-09 13:31 · Score: 1

Thad,

Itanium and Hammer aren't yet suitable for an investment as large as Pixar has just made. The chips and motherboards aren't cost-effective compared to 32-bit, and there's not enough experience with that hardware yet to point out all of the gotchas. Next time. They'll probably do another render farm in a year. They want to get 3 or 4 times their present size - at least in square feet of office space. Crazy as that sounds, if you've seen the present huge campus.

Bruce

--
Bruce Perens.

Re:Why not go with Itanium or the new Hammer? by MadFarmAnimalz · 2003-02-09 22:23 · Score: 1

Is that expanding office space, expanding processing power in general, or maxing out processing power per unit of office space?

--
Blearf. Blearf, I say.

A higher moderator? by Idou · 2003-02-09 14:08 · Score: 1

That's kind of funny. My post went from 5 insightful to redundant, and now I can't see how many mods it got . . . makes me wonder about the credibility of slashdot moderation.

Or does this just mean that my point is "moot" and therefore redundant?

--
Sdelat' Ameriku velikoy Snova!

Re:A higher moderator? by sirsnork · 2003-02-09 15:47 · Score: 1

makes me wonder about the credibility of slashdot moderation.

HAHAHAHAHAHAHA.

That's the funniest thing I've read on /. for a LONG time

--

Normal people worry me!

Re:This is exactly the kind of thing x86 is good a by Quarters · 2003-02-09 14:08 · Score: 1

Sun is going to be around for a long time. As many other people have pointed out, they're just retreating somewhat to more a of niche market, where they are the right tool for the job.

Sun has always served a niche market. The problem is that commodity X86 hardware, coupled with a decent OS, is making their niche smaller and smaller. Eventually commodity hardware/software will be capable of doing everything in that niche.

Sun needs to find ways to move into other markets. Retreating is not the strategy of a company that will "be around for a long time".

Oracle!! by NineNine · 2003-02-09 14:17 · Score: 1

One place that I *always* see Sun boxes are Oracle instances. All of the big (3+ terabyte of data) instances I've seen are always running on Sun. I'm guessing that Oracle's version of their products on Sun are the most stable & most efficient. Hell, if things got bad enough, Sun could team up with Oracle and make dedicated boxes. I think that they'd do very well.

Re:Why not Apple? by be-fan · 2003-02-09 14:33 · Score: 1

Only if the data set fits in the L3 cache of the G4. The G4 extremely limited memory bandwidth, so while the AltiVec unit is theoretically 2-3 times as fast at a given clock speed (which makes it comparable to maybe 1.5 Xeons overall) it's totally demolished by the meager 1.3 GB/sec of main memory bandwidth. Data sets like these are extremely bandwidth limited. You know, SSE stands for "Streaming" SIMD Extensions for a reason.

--
A deep unwavering belief is a sure sign you're missing something...

Steve Jobs + Pixar by Biggs+Driut · 2003-02-09 14:58 · Score: 1

Is Pixar still owned by Steve Jobs? Shouldn't they be using xserve? how 'bout some NEXT servers? why not an Apple ][ ? (just kidding on that last one....) Not that the servers they are using now are bad, its just that there has to be a reason Steve's company isnt using his computers. Just a thought.

Vacum Tube's VS Mosfet's by iamagod420 · 2003-02-09 15:01 · Score: 1

I'll bet that people will claim that they can tell when a scene has been rendered with SUN, and when a scene is rendered with Intel. Just like the ass's that claim they can tell the difference between a song played through a Tube Amp, and one in a modern one. Modern sounds better... just as intel rocks.

RackSAVER, not RackSPACE by killmode · 2003-02-09 18:20 · Score: 1

one is a High Density Clustering Solutions Company,
one is a Webhosting Company.

News.com got it wrong earlier tonight, but they have fixed the link.

Http://www.racksaver.com/

fyi Most of the best movies rendering CGI are RackSaver Cluster customers.

check them out amazing GFLOPS per $.

Killmode

And Dell does essentially JACK. by Ayanami+Rei · 2003-02-09 18:46 · Score: 1

They farm out all of their R&D in the form of browsing the latest offerings from Taiwanese system integrators. That R&D budget buys hookers and pays engineers who muddle the power connectors on each new motherboard.

So saying Sun spends 10 times more than that ain't saying much.

--
THIS THING CAN TURN ON A DIME, MACROSSZERO STYLE ALSO FUCK BETA, ~NYORON

Sun seeks self-reliance with new gear by girish · 2003-02-09 21:40 · Score: 1

Sun Microsystems will overhaul most its server line and cut prices Monday in an effort to show that it's got enough research power to stay competitive, even with today's depressed spending.

http://rss.com.com/2100-1001-983913.html?type=pt&p art=rss&tag=feed&subj=news

Airbus Computers by Puu · 2003-02-10 02:40 · Score: 1

I don't remember it exactly or have a link handy... But I remember the computer system in the early Airbus aircraft consisted of four machines, each with two halves. While each of the four machines had specialised tasks, each was nevertheless capable of doing the others' tasks in case of malfunction; each computer ran different code for the same task. Furthermore, the two halves of each computer ran different code and the halves constantly monitored each other -- if results differed, another computer took over the failing task. So there was pretty good error checking and system redundancy.

Why the early Airbus planes do lead the disaster charts has mainly to do with pilot interface. It was possible to accidentally have the aircraft in "Cruise mode" or "Takeoff mode" at landing, and there was no loud and clear warning system -- this led to some crashes. Another related issue was the abundance of numerical information in place of the traditional round gauges -- you didn't notice a problem so quickly (at a glance), if at all. [Again no linkage, I read this in a book on aircraft disasters.] Due to poor interface design, the aircraft were very prone to human error, despite all the safeguarding automation.

Moot point, though. Obviously Boeing won't employ WinXP in the aircraft but firmly on the ground.

Re:Raw CPU power- Exactly! by pmz · 2003-02-10 03:07 · Score: 1

The IT people I talked to were surprisingly happy with XP so far.

What is so sad about this is that Windows XP still is nowhere as robust as Solaris 8 or 9. When I use Windows XP and 2000, they still feel very much like Microsoft made them (inconsistent behaviors, applications can still crash them, etc.). On the other hand, I beat the hell out of Solaris 8 at work (CAD development), and it takes it without a whimper. My only downtime is scheduled maintenance. I can tell from experience that Sun follows better software engineering procedures than Microsoft (at least for their OS kernel). Also, even Linux is better than Win XP and Win 2K (I used Linux extensively while in college and was very pleased).

The only advantages to Windows are the Microsoft sales pitch and Office. Both have limited days ahead of them...mainly due to Linux and OpenOffice.

--
Healthcare article at Kuro5hin

Raytracing Available by Puu · 2003-02-10 03:57 · Score: 1

However, their site says RenderMan now offers raytracing as an additional tool for effects. (While it's still a scanline renderer.)

Re:Why not Apple? by Puu · 2003-02-10 04:19 · Score: 1

Why not Apple?

Heck, why not 1024 iMacs!

Imagine a lovely field of sunflowers.

Re:This is exactly the kind of thing x86 is good a by jafac · 2003-02-10 07:17 · Score: 1

What it REALLY means is that Apple is dying.

I mean, with SJ as CEO of both Pixar AND Apple, you'd think he'd buy up a buttload of X-Serves, and run Darwin. . .

--

These are my friends, See how they glisten. See this one shine, how he smiles in the light.

All Your Base Are... by Puu · 2003-02-10 07:18 · Score: 1

... e?

Re:1024 processers by Puu · 2003-02-10 07:26 · Score: 1

Also look up the "KLAT" cluster supersomputer. It has a very cleverly routed, efficient network, with only three (IIRC) NICs per box.

They understood not everything has to connect to everything else. Wonder if their method is doable on a rack blade server?

RackSaver's BladeRack by jimmyfp · 2003-02-10 14:02 · Score: 1

The news about Pixar should say that Pixar purchased 8 BladeRacks from RackSaver, not RackSpace. Some moron editor decide to re-write the original press and didn't do a very good job. Anyone that wants the correct and original press should go to www.racksaver.com to see the press. The first 25 people that reply to this post can get a free RackSaver T-shirts. Or email me at jimmy@racksaver.com.

Re:I wonder... by mrm677 · 2003-02-11 05:02 · Score: 1

They have a cache of nearly every internet document. Even 4gb of memory in each of those 15,000 boxes is 60 terabytes. Maybe that's enough, but I doubt it because they _never_ throw away stuff from the cache. The Google engineer actually addressed this issue because they have a lot of "power". For example, they can look at how websites change over time if they wanted to.

Also, they actually use the filesystem for communication between nodes. Messages are written to a log on some type of distributed file system.

Slashdot Mirror

Pixar Eclipses Sun with Linux/Intel

312 of 437 comments (clear)