Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Pixar's Render Farm

twitter.com

Post reply on HN
Print view View original

241–250 of 323 posts

Re: Pixar's Render Farm

#241

Post by shaklee3 »

berkut wrote: post #125
dahart wrote: post #117

Earlier quoted context omitted.

> Because the expense is not really worth it I disagree with this takeaway. But full disclosure I’m biased: I work on OptiX. There is a reason Pixar and Arnold and Vray and most other major industry renderers are moving to the GPU, because the trends are clear and because it has recently become ‘worth it’. Many renderers are reporting factors of 2-10 for production scale scene rendering. (Here’s a good example: https…

I write a production renderer for a living :) So I'm well aware of the trade offs. As I mentioned, for lookdev and small scenes, GPUs do make sense currently (if you're willing the pay the penalty of getting code to work on both CPU and GPU, and GPU dev is not exactly trivial in terms of debugging / building compared to CPU dev). But until GPUs exist with > 64 GB RAM, for rendering large scale scenes, it's just not w…

GPUs do exist with 64+GB of rao, virtually. A dgx2 has distributed memory where you can see the entire 16x32GB of address space backed by nvlink. And that technology is now 3 years old, and it's even higher now.

Re: Pixar's Render Farm

#242

Post by shaklee3 »

boulos wrote: post #231
dahart wrote: post #117

Earlier quoted context omitted.

> Because the expense is not really worth it I disagree with this takeaway. But full disclosure I’m biased: I work on OptiX. There is a reason Pixar and Arnold and Vray and most other major industry renderers are moving to the GPU, because the trends are clear and because it has recently become ‘worth it’. Many renderers are reporting factors of 2-10 for production scale scene rendering. (Here’s a good example: https…

Dave, doesn’t that video show more like “50% faster”? Here’s the timecode (&t=360) [1] for the “production difficulty” result (which really doesn’t seem to be, but whatever). Isn’t there a better Vray or Arnold comparison somewhere? As in my summary comment, an A100 can now run real scenes, but will cost you ~$10k per card. For $10k, you get a lot more threads from AMD. [1] https://m.youtube.com/watch?v=ZlmRuR5MKmU&t…

What do you mean by a lot more threads? Are you comparing an epyc?

Re: Pixar's Render Farm

#243

Post by Razengan »

Twitter really should just add support for long ass tweets instead of this janky mess of trying to pick out all the crumbs of people replying to themselves.

How the hell can we have so much tech and leave so many common use cases so fucking broken for so long

It’s a perfect example of how you don’t really decide how others use your products; people do.

Re: Pixar's Render Farm

#244

Post by blhack »

nikolay wrote: post #47
hadrien01 wrote: post #10

For those that can't stand Twitter's UI: https://threadreaderapp.com/thread/1345146328058269696.html

It's very painful to follow a conversation on Twitter. I'm not sure why they think the way they've done things makes sense.

It makes people misunderstand each other, which makes them angry, and increases engagement/replies.

Re: Pixar's Render Farm

#245

Post by alasdair_ »

Around 2002 I worked on Sun Microsystem’s Grid Engine and their “go to” demo for new customers was always a big Pixar render farm job. There was something about seeing the final rendered clip that convinced people that massively parallel work was the way to go for their use case.

Re: Pixar's Render Farm

#246

Post by boulos »

shaklee3 wrote: post #242
boulos wrote: post #231

Earlier quoted context omitted.

Dave, doesn’t that video show more like “50% faster”? Here’s the timecode (&t=360) [1] for the “production difficulty” result (which really doesn’t seem to be, but whatever). Isn’t there a better Vray or Arnold comparison somewhere? As in my summary comment, an A100 can now run real scenes, but will cost you ~$10k per card. For $10k, you get a lot more threads from AMD. [1] https://m.youtube.com/watch?v=ZlmRuR5MKmU&t…

What do you mean by a lot more threads? Are you comparing an epyc?

Yeah, that came off clumsily (I’d lost part of my comment while switching tabs on my phone).

An AMD Rome/Milan part will give you 256 decent threads on a 2S box with a ton of RAM for say $20-25k at list price (e.g., a Dell power edge without any of their premium support or lots of flash). By comparison, the list price of just an A100 is $15k (and you still need a server to drive the thing).

So for shops shoving these into a data center they still need to do a cost/benefit tradeoff of “how much faster is this for our shows, can anyone else make use of it, how much power do these draw...”. If anything, the note about more and more software using CUDA is probably as important as “ray tracing is now sufficiently faster” since the lack of reuse has held them back (similar things for video encoding historically: if you’ve got a lot of cpus around, it was historically hard to beat for $/transcode).

Re: Pixar's Render Farm

#247

Post by dahart »

boulos wrote: post #231
dahart wrote: post #117

Earlier quoted context omitted.

> Because the expense is not really worth it I disagree with this takeaway. But full disclosure I’m biased: I work on OptiX. There is a reason Pixar and Arnold and Vray and most other major industry renderers are moving to the GPU, because the trends are clear and because it has recently become ‘worth it’. Many renderers are reporting factors of 2-10 for production scale scene rendering. (Here’s a good example: https…

Dave, doesn’t that video show more like “50% faster”? Here’s the timecode (&t=360) [1] for the “production difficulty” result (which really doesn’t seem to be, but whatever). Isn’t there a better Vray or Arnold comparison somewhere? As in my summary comment, an A100 can now run real scenes, but will cost you ~$10k per card. For $10k, you get a lot more threads from AMD. [1] https://m.youtube.com/watch?v=ZlmRuR5MKmU&t…

Yes, this example isn’t quite as high as the 2-10x range I claimed, but I still liked it as an example because the CPU is very beefy, and it’s newer and roughly the same list price as the GPU being compared. I like that they compare power consumption too, and ultimately the GPU comes out well ahead. There are lots of other comparisons that show huge x-factors, this one seemed less likely to get called out for cherry picking, and @berkut’s critique of texture memory consumption for large production scenes is fair... we’re not all the way there yet. But, 50% faster is still “worth it”. In the video, Sam mentions that if you compare lower end components on both sides, the x-factor will be higher.

Re: Pixar's Render Farm

#248

Post by a_e_k »

mroche wrote: post #196
dgrant wrote: post #175

Earlier quoted context omitted.

Hi. I used to work at Teradici. It was always interesting that Pixar went with VDI because it meant the CPUs that were being used as desktops during the day could be used for rendering at night. Roughly speaking. The economics made a lot of sense. A guy from Pixar came to Teradici and gave a talk all about it. Amazing stuff. Interesting contrast with other companies that switched to VDI where it made very little sens…

Yep, all of their workstations were dual socket servers, where each socket was a workstation VM with PCIe passthrough, and each getting their own hostcard+GPU. Each VM had dedicated memory, but no ownership of the cores they were pinned to, so overnight if the 'workstations' were idle, another VM (also with dedicated memory) would spin up (the other VMs would be backgrounded) and consume the available cores and add i…

That's one reason for the checkpoint feature, yes, but there are others. A few years back (Dory-era), I participated in a talk at SIGGRAPH '15 about some of them:

https://dl.acm.org/doi/abs/10.1145/2775280.2792573

http://eastfarthing.com/publications/checkpoint.pdf

Re: Pixar's Render Farm

#249

Post by ganafagol »

CamJN wrote: post #233
jlgaddis wrote: post #229

Earlier quoted context omitted.

If the very light gray color of your comment is any indication, folks here didn't like what you said. I certainly don't think that HN should ban you because of it, though.

[flagged]

Reinforcing the walls of your bubble doesn't make it less of a bubble. No matter left or right.

Re: Pixar's Render Farm

#250

Post by ganafagol »

nikolay wrote: post #205
yiyus wrote: post #134

Earlier quoted context omitted.

It may be difficult to evolve, but it should be possible.

If there's will, there's a way. Remember how long they've tried to convince themselves that 140 characters is more than enough. Nowadays, even phones display conversations better and combine or split text messages hiding the 160 characters limit from the end user and allow longer than 280 characters texts. Twitter has some nice ideas, but it's mostly a huge missed opportunity.

The character limit was central to making Twitter what it is. Without that limit you'd just have ended up with walls of text and nobody would have been interested. This is a great example of limitations forcing creativity.

The character limit is the defining feature of Twitter.

Post reply on HN