Rent A Bit Of Weta Digital

An anonymous reader writes linking to this story at stuff.co.nz, excerpting: "Five hundred powerful computers used by Weta Digital to help create the special effects for the Lord of the Rings may be put up for hire.... The pizza-box sized IBM blade servers each incorporate dual 2.8 gigahertz Intel Xeon processors and 6 [gigabytes?] of memory." Update: 03/22 07:08 GMT by S : The linked story says 6 megabytes of memory, we don't believe 'em.

Re:Distributed.net...

[SillySnake]

A great idea in theory, but how would they track the amount of help that you did in a way that would be one hundred percent hack proof? I don't think you'd want to pay people to analyze every packet that you get back to make sure it's had whatever needed done, done to it. Granted, it would be possible to elimnate most unwanted results with a couple of filters, but when money becomes an issue the community will do what they can to get the most of it.

I don't see why not.

[torpor]

If you've got renderman set up to render to disk, and your disk arrays are pretty fast, I don't see any reason why these dedicated render machines shouldn't have only 6 megabytes of RAM per CPU.

okay, it doesn't make a -ton- of sense to render direct to disk, but maybe it can be done and not require so much RAM?

interconnect

[painehope]

the real killer is that there's quite a few industries that can't rent time on their cluster because the gigabit interconnect ( IBM blade chassis have a switch module internal to each chassis, and I don't think you can get any HSLL - high-speed, low-latency - network interconnect modules ( Myrinet, SCI, Quadrics, etc. ) for them ) has too high of a latency for their applications.

Bandwidth-wise they should be fine, as each chassis has at least four ports that could be trunked to a top-level switch w/ a beefy backplane ( I could tell you the # of ports per chassis if I was at work, as I've been messing w/ some of their blades lately ), giving a peak per-chassis bw of > 400 MB/sec.

Of course, I'm wondering how Weta got around it themselves, as I would think that rendering digital video is fairly heavy on inter-node communication. This would still be aswesome for web-servers or problems that are "embarassingly parallel".

Re:interconnect

[2megs]

Rendering digital video is about as parallel as compute loads get. Generally each frame can be an independent computation. For most ray-tracing algorithms, computing each pixel of each frame is fully parallelizable too.

The global AI things they did to have 10,000 troops all interacting together is obviously not quite so independent, but I'm willing to be the bulk of the compute load goes into creating pictures of those interactions, not the interactions themselves.

Re:LAN Connection ?

[Anonymous Coward]

That joke is so old that people get it even when you misspell the punch line.

Re:I don't see why not.

[Jexx Dragon]

You said it right there... 6MB of RAM per CPU... Five hundred computers... Thats a mighty big number. Since its a cluster each node has to render only a small part of the image and can render to disk.

Re:One thing to say...

[CGP314]

Using your backassward logic, it seems more logical to devote your CPU time to researching automotive traffic patterns, so you don't get killed in an auto accident or get hit by a bus.

If there was a project that I could devote my CPU cycles that could reduce the possibility of me getting into a car accident, then I would drop folding@home for dontgethitbyacar@home. What's backassward about risk assesment?


-Colin [colingregorypalmer.net]

Re:Renderfarms online - old news

[rimu guy]

Boy did WETA fuck up on this one.

Boy, did they ever. The graphics in the Lotr and Master and commander sucked. The movies were released late. No one won any awards. And the NZ film industry is in tatters.

No wait. I mean the opposite of all that.

Weta knew up front the boxes were only of any use to them for a couple of years. They _budgeted_ on throwing them away after the Lotr trilogy was done. If they can get anything back on them now, more power to them.

Personally, I think its very cool. I'm even seeing if I can get a couple of machines from them to host some of my Lotr-fan customers on. I for one would be keen to run on an ex-Weta server.

- Linux VPS Hosting [rimuhosting.com]

Re:interconnect

[Obasan]

Actually, rendering is fairly light on network requirements and very heavy on memory/cpu. (Download scene files & textures then crunch numbers for 10-40 minutes depending on layer complexity.)

But the bladecenter chassis also does in fact support a Myrinet interconnect if you so desire.

Now we find out if "grid computing" sells

[Animats]

OK, "grid computing" fans, here it is, a big CPU resource open for commercial customers. Let's see if people line up to buy cycles. There must be paying customers out there who want to do rendering, or VLSI simulation, or numerical wind-tunnel tests of wing sections, or something.

We're waiting...

As I've pointed out before, if there was a market for this, ISPs would be selling off-peak CPU time on their hosting farms.