How Pixar Is Reinventing Rendering with XPU

Summary

---

Podcast Summary

Podcast: Reshaping Workflows with Dell Pro Precision & NVIDIA RTX PRO GPUs

Episode: How Pixar Is Reinventing Rendering with XPU
Host: Logan Mahler (Dell Technologies)
Guests: Dylan Sisson (Pixar), Cindy Oli (Dell Technologies)
Release Date: March 31, 2026

---

Episode Overview

This episode dives into the groundbreaking evolution of Pixar’s rendering pipelines, focusing on the introduction of RenderMan XPU—a new hybrid CPU/GPU renderer powering the visuals for Toy Story 5. Dylan Sisson, veteran at Pixar and Head of RenderMan Creative Engagement, discusses Pixar’s four-decade legacy of innovation, technical breakthroughs, and the real-world impact of embracing new hardware in partnership with Dell and NVIDIA. The conversation covers Pixar’s technological culture, the shifting landscape of rendering, tangible differences XPU brings to animation, and the wider creative community through events like SIGGRAPH.

---

Key Discussion Points & Insights

1. Pixar’s 40th Anniversary and RenderMan Origins

[02:04–05:26]

• Dylan Sisson provides a historical overview, noting “Today is Pixar’s 40th birthday, February 5th” and sharing anecdotes about Pixar’s roots as a commercial and tech studio.
• RenderMan’s development stemmed from a need to create imagery for film that existing hardware couldn’t support; Pixar designed its own software and even hardware.
• Major technical legacies:
  • Developed the first “shading language” and even the ubiquitous alpha channel (“You can’t imagine a world without an alpha channel. It’s like imagining a world without pizza. It’s impossible.” — Dylan Sisson, 04:21).
  • Successive inflection points: shift from efficient scanline rendering to ray tracing and Monte Carlo path tracing.

2. Introducing RenderMan XPU: The Hybrid Revolution

[05:46–08:10]

• XPU designed as a renderer that leverages both CPU and GPU resources—"independently, or both at the same time."
• The main philosophy: get artists closer to “final pixels” faster, increase creative interactivity, and collapse technical barriers between departments.
• “If you’re an artist…you can bring all that hardware to bear…our focus is providing interactivity for artists…allowing people to make more decisions sooner, eroding friction points in the pipeline.” — Dylan Sisson, 06:25

3. Technical & Creative Breakthroughs (and Challenges)

[08:56–10:06]

• Key foundational decision: XPU must produce identical pixels regardless of hardware—“We didn’t want to have one renderer for the CPU and one for the GPU. We wanted one renderer that abstracted…hardware. So we can…get the same pixel, which is extremely hard.” — Dylan Sisson, 09:13
• XPU’s debut on Toy Story 5 allowed the entire pipeline (shaders, lights, assets) to migrate seamlessly, a milestone not possible in previous renderer transitions.

4. Rendering Evolution: From Toy Story to Toy Story 5

[10:10–11:12]

• Shot realism, stylization, and the ever-growing ambition of visuals are discussed.
• “We’re always trying to surprise the audience with new kinds of images…Toy Story 5 is packed with that kind of stuff.” — Dylan Sisson, 10:32
• Playful teases: “The Pixar walking teapot might make a small cameo in Toy Story 5.” — Dylan Sisson, 10:58

5. AI, Machine Learning & Denoising in Production

[14:33–16:46]

• Dylan explains Pixar’s practical use of “narrow AI”—especially denoising with machine learning, first used on Toy Story 4.
• Machine learning denoisers cut rendering samples (and cost), enabling imagery that would have previously been impossible.
• “There’s people at Pixar who’ve worked in production a long time…they call the denoiser ‘magic.’ It just transforms what they can do and when they can go home.” — Dylan Sisson, 16:21

6. XPU in Workflow: Testing, Deployment, and Hybrid Compute

[16:53–19:05]

• XPU abstracts hardware: “You can basically pick which one [CPU or GPU]…jobs are allocated based on operation type."
• Robust, multi-environment testing at Pixar, including on a variety of Dell and other workstations.
• Live production (“best beta tester”) is essential for stress-testing new rendering technology.

7. Render Farms: On-Premises vs. Cloud

[19:05–21:14]

• Pixar’s hybrid approach: TDs (Technical Directors) check out servers by day (local & VM), then those become render farm nodes at night.
• Some cloud bursting for large workloads.
• Blinn’s Law: No matter how fast computers get, render times for movie frames stay constant due to increasing creative ambition.
• Empowering both studios and individuals to harness all hardware locally—or offload to the cloud as required.

---

Community Engagement & Pixar Traditions

1. The RenderMan Art & Science Fair at SIGGRAPH

[21:38–24:06]

• An annual community event at SIGGRAPH: networking, demos, and the legendary collectible “walking teapot” giveaway.
• “We have…our signature drinks, including the Render Manhattan…and at the end, everybody gets a walking teapot.” — Dylan Sisson, 21:54
• Began in 2003 as a whimsical project that “was actually cheaper than printing brochures.” Now a coveted prize—with some teapots selling for over $600 on eBay!

2. The RenderMan Challenge

[25:11–28:06]

• Now in its 13th year, invites artists worldwide to tell a visual story in a single image using provided assets and RenderMan (non-commercial/XPU encouraged).
• Huge community engagement, mentorships, and job opportunities—“A platform…that winners have been able to put on their resume and attribute them getting jobs to having participated.” — Dylan Sisson, 26:37
• Upcoming challenge launches at SIGGRAPH July 2026; open to all skill levels using free tools.

---

Notable Quotes & Memorable Moments

• “Imagine a world without an alpha channel…It’s like trying to imagine a world without pizza. It’s impossible.” — Dylan Sisson, 04:21
• “We can render pretty much whatever we want. But the challenge really is about enabling creative flow, allowing people to make more decisions sooner.” — Dylan Sisson, 06:43
• “We wanted to have one renderer that abstracted the ideas of the hardware…get the same pixel, which is…extremely hard.” — Dylan Sisson, 09:13
• “There’s people at Pixar…not impressed by anything…they call the denoiser ‘magic.’” — Dylan Sisson, 16:21
• “The teapot became very popular…seen them go on eBay for over $600, which is more than a license of RenderMan.” — Dylan Sisson, 24:41
• “It’s just a really fun way to explore storytelling in a single image with…cutting edge hardware. That’s RenderMan XPU…the same we’re running at the studio.” — Dylan Sisson, 27:37

---

Useful Timestamps

• 02:04 — Pixar’s founding & technology origins
• 05:46 — Core philosophy behind RenderMan XPU
• 08:56 — Technical challenges in making XPU “final frame” ready
• 10:32 — What audiences will notice in Toy Story 5
• 14:33 — Practical uses of AI/machine learning in Pixar’s pipeline
• 16:53 — How XPU distributes CPU vs GPU workload
• 19:05 — Pixar’s hybrid approach: local vs. cloud rendering
• 21:38 — SIGGRAPH’s Art & Science Fair and the walking teapot
• 25:11 — RenderMan Challenge: origins, prizes, and community impact
• 28:22 — Where to find Dylan Sisson online & how to download non-commercial RenderMan

---

Further Resources

• Try RenderMan XPU: Download the non-commercial version (with no watermarks/limitations) at renderman.pixar.com
• Connect with Dylan Sisson: LinkedIn & Instagram (@realDylanSisson)
• Learn, Compete, & Network: Join the RenderMan Challenge and SIGGRAPH Art & Science Fair annually – watch for dates via Pixar’s site and Discord.

---

This episode is a must-listen for anyone interested in the intersection of creativity and cutting-edge tech—from production artists and students to pipeline engineers. The discussion sheds light on how hardware advances (Dell, NVIDIA) and software breakthroughs empower the movie magic of today’s—and tomorrow’s—Pixar films.

Transcript

[Dylan Sisson]

Foreign.

[Cindy Oli]

Welcome to Reshaping Workflows with Dell Pro Precision and Nvidia, where innovation meets real world impact in high performance computing.

[Logan Mahler]

Welcome back to another riveting episode of Reshaping Workflows with Delpro Precision, Nvidia RTX 2000 fused. I'm your host, Logan Mahler. Today, big name on the podcast, very excited, never had this person on before. We have Dylan from Pixar who's joined us to talk about a few things, gtc, a few things XPU and all things Pixar. But before I do that, because I want to save the best for last, I'm going to go ahead and kick it to my partner in crime, my better work half. Cindy, you haven't been on a podcast in a while. Please introduce yourself so everyone remembers you.

[Cindy Oli]

Maybe because you don't invite me to them. Hi, everyone, Cindy Oli, program manager for all things media and entertainment and excited to be here.

[Logan Mahler]

Wouldn't be the same without you. Now, moving on, Dylan, tell everyone thank you for joining. Obviously, tell everyone a little bit about you introduction, background, what you do, and then we'll jump right into it.

[Dylan Sisson]

My name is Dylan Sisson. I started at Pixar in 1999, so in the last, last century. And I've been been in the Renderman group since then in a number of different roles. Now I'm the head of Renderman Creative Engagement. I do a lot of things, from artistic R and D to educational outreach, to going around talking about all the good work that we're doing at Pixar to make RenderMan of GPU accelerated with RenderMan XPU. I think you have some cool stuff we're talking about this year, Dylan.

[Cindy Oli]

I'm going to start at the very beginning, like you guys. You know, Pixar started off as an animation studio. What made them want to develop their own proprietary software? Like what was the genesis of Renderman?

[Dylan Sisson]

This is a big question. Let's go back to the founding of Pixar in 1986. And you both might be surprised to know that today is actually. Today is Pixar's 40th birthday, February 5th. As of this recording, that's it. We have balloons all around the studio. If you wear a Pixar shirt, you get a free lunch in the cafeteria. It's a big birthday. So about 4,0. And when Pixar was founded, people think about Pixar as a movie studio, but the technology at the time, when Pixar was founded, it would cost billions of dollars to have bought the hardware to even Render a movie. So they used Moore's Law, exponential growth and calculated they could make a movie in 1995. And that's what they planned to do. And they made Toy Story. Up till that point, Pixar was creating the Pixar image computer. You might have heard about hardware. We created 72 commercials for TV. So Pixar originally was like a commercial studio and we were rendering everything with our developing visual effects animation pipeline. And Renderman was the core of that. It was a very efficient scanline renderer that was able to render things that a ray tracer couldn't at the time. And the core algorithm, the Rays algorithm, actually predates Pixar and came from the computer Graphics group with all the good work that Ed Catmull, Alvirae Smith, Lauren Carpenter, all those guys solved all the problems in the world that needed solving to render visual effects and animation, including solving things like jaggy edges. They wrote a shading language so you could express all these different arbitrary things in it. And you may have heard of the alpha channel, like rgba. They invented the alpha channel. So imagine a world without an alpha channel. You can't. It's like trying to imagine a world without pizza. It's impossible. So Renderman really was this idea of being able to render high fidelity computer graphics when everything people were rendering was just kind of looked more like toyish, but something for the feature film screen. So we came out with Toy story in 1995 with the Renderman as the Raze renderer. We had a couple inflection points when we added ray tracing to it. Machines got faster, we got more memory, we could start ray tracing these data sets that were previously impossible. Then we had the RAS architecture for finding gory, which allowed us to bring like full Monte Carlo path tracing to bear on these production scenes, which up to that point we couldn't do. And that's the renderer that we have today. This last year has been notable because we've been able to roll out a brand new architecture. So it's not a new feature, but it's a new renderer. And that renderer is Readyman xpu. And we're finally using it to render final frames on Toy Story 5, which is historically significant. Previously, even two years ago, we wouldn't have been able to render like a full production Pixar scene on a gpu. Today we can. With the new Blackwells and the adas, it's really interesting.

[Cindy Oli]

I love that Pixar is a geriatric millennial. I understand the feeling, but I Love that. With this new xpu, you guys are combining the CPU and the gpu. Can you explain what the core philosophy around XPU is and what that's going to mean to artists who may still be using CPU only workflows?

[Dylan Sisson]

Yeah, I think at a certain time in computer graphics, it was a struggle to do anything right. So when we rendered like Toy story in 1995, there was a reason why we made a movie about plastic toys, and that's because everything we rendered back then looked like plastic. So a lot of it was overcoming technical challenges to get these images up on screen. Today, we can. With light transfer and the amount of compute and memory we have, we can simulate what light's actually doing, and we can make skin look like skin, metal look like metal, and all that kind of amazing stuff. So the goal of XPU was, when we started working on it, we projected in the future and said, hey, at a certain point, we're going to have cards that can render a Pixar scene, and we should build a renderer that can handle that. So XPU is able to use the cpu, the gpu independently, or both at the same time. And if you're an artist sitting and you have a workstation that has a good GPU and good CPUs, you can bring all that hardware to bear on whatever you're rendering. And our focus today is providing, like, interactivity for artists, so we can make time to first decision faster on these very complex shading networks that we have on our assets. So that's really what's exciting about it, is we can render pretty much whatever we want. But the challenge really is about enabling creative flow, allowing people to make more decisions sooner, kind of eroding these friction points in the pipeline between layout and animation simulation allow people to render final pixels much earlier in production than they could before. So it's. It's actually pretty exciting and fun. And as an artist who uses and plays around with it, I'm like, this is cool. You know, it's. It's actually fun. Fun to work with.

[Cindy Oli]

I love thinking back on it. I can't believe Toy Story, did you say? Came out in 95. It was. I mean, growing up in the 90s, it was all of the, you know, the Disney 2D sort of Renaissance Disney stuff, you know, the Beauty and the Beast and the Little Mermaid. And then Toy Story came out and it was like this light switch where it was like, oh, my gosh, what is this? And then everything after that was different. So, I mean, I feel like you guys were that defining moment where animation completely changed and, yeah, it's always been so exciting. And we're big fans of Pixar in this house, so I love it.

[Dylan Sisson]

Yeah, that's cool. Yeah, the. The whole digital effects revolution, like with Toy Story and also with Jurassic park, like, ILM used RenderMan to render the dinosaurs that people saw the movie and they're like, what am I seeing? Like, these can't be real dinosaurs. How did they do it? Like, during those five years, there really was a revolution in digital visual effects that we're still enjoying today.

[Cindy Oli]

I know. I mean, I look back on these movies and, like, this is pretty good for the time. I mean, things have obviously gotten better. But, like, watching that as, like a kid and being like, oh, my gosh, these are real dinosaurs, it was. It was mind blowing. So what were the biggest technical or creative challenges that, you know, the Pixar team faced in making XPU final frame ready?

[Dylan Sisson]

We wanted to do it right, and that meant making all the correct foundational decisions. We didn't want to do a bunch of hacky stuff. And one of the early decisions that we made was that we didn't want to have one renderer for the CPU and one for the gpu. We wanted to have one renderer that abstracted the ideas of the hardware. So we can essentially, with xpu, render on the cpu, the GPU and get the same pixel, which is something that's extremely hard to do, and we spend a lot of time thinking about that. And our current renderer, our Monte Carlo renderer, RAS produces the same pixels as well. So for Toy Story 5, we were able to swap out RAS with XPU using the same shaders, same lights, same assets, and get the same image out, which is kind of. Kind of remarkable. When we did the switch from rays to ras, we had to have completely different shaders, completely different lights. That's one of the, probably the biggest fundamental difference or challenge with making XPU

[Cindy Oli]

XPU and looking at, like, Toy Story 4 and then Toy Story 5, you know, when I'm in the theater, what are the shots that I'm going to see that I'm really going to be able to see? The dramatic differences between, you know, Rennerman 4 and Renderman 5.

[Dylan Sisson]

I can't really spoil anything but shots

[Cindy Oli]

or scenes or like, anything of, like, is it something with lighting, is it movement? Like, what are. What am I going to be looking for?

[Dylan Sisson]

I think we can say that the. We're always trying to surprise the audience with new kinds of images. We haven't seen before. And Toy Story 5's packed with that kind of stuff. I think there's some interesting things from the stylization standpoint that might surprise people. I can't say anything but the realism of Toy Story 4. Like, we want to keep within the same world of Toy Story, but also enhance it and surprise it. So there's a Be excited to see what's happening. And I can't confirm or deny this, but the Pixar walking teapot might make a small cameo in Toy Story 5.

[Cindy Oli]

Can you tell us, will there be an opening short at the very least?

[Dylan Sisson]

You know, with the. The shorts have been a brave, fun tradition that Pixar's been doing. We'll see what we get in front of Toy Story 5. I'm not sure if we've nailed that down yet.

[Cindy Oli]

It's one of my favorite parts of any Pixar movie. It's the opening short. I love them.

[Dylan Sisson]

I love the shorts, too. What's your favorite short?

[Cindy Oli]

Oh, Sandpiper is up there. That one was one of my favorites.

[Dylan Sisson]

Me too.

[Cindy Oli]

I loved it. It was just so cute. That was one of my favorites. The bao one was adorable. Yeah. There's just been so many.

[Dylan Sisson]

Bao was awesome.

[Logan Mahler]

This is going to show that I know nothing about media entertainment. But didn't you do one with a stork? Like the baby and the stork?

[Dylan Sisson]

We did one called for the Birds, where the stork lands on the power line and all the birds kind of like all the little birds kind of slide into him. Remember that one?

[Cindy Oli]

He doesn't remember that one. Remember yesterday? What are you talking about?

[Logan Mahler]

Hey, man, it's probably true. I'm losing my mind, but I could swear there was a short of like, something with a stork or something.

[Dylan Sisson]

Yeah, I think. I think you. You lucked out there before the birds. Yeah.

[Cindy Oli]

For the record, your favorite Pixar movie.

[Dylan Sisson]

My favorite Pixar movie. I get this asked this once in a while. I think my favorite out of the top top is ratatouille. I like technically rendering beautiful any movie that makes me want to eat food. That's cg. I'm like, wow, that's amazing. Technically, there's so much cool stuff going on with how we used point clouds to do A Radiance. And I just loved all the furry, furry rats. And, you know, Jan's idea of putting the rat underneath the chef's hat I thought was just super brilliant and something fun to commit to. You know, it's one of those inflection points where, you know, if you look at Monsters Incorporated at the time, people were like, this is a great movie. Did you see solely like how hairy he was? And nobody really was like, why do you only have one furry monster in your movie? And that's because we could only render one at a time. Like we had a couple monsters with fur, but most of them had scales or hides or something that. That didn't require fur. That was a limitation at the time for Ratatouille. We could render as many furry things as we wanted, which was just super awesome.

[Cindy Oli]

That's awesome. I love seeing your name in the credits. I'm always looking for them too. It's like the one guy that I know that's in major movie credits. So that's.

[Dylan Sisson]

Yeah, it's at the very end, of course.

[Cindy Oli]

Hey, they're in there. That's what matters in there. So, Mr. I over here, Logan Lawler, do you have any questions?

[Logan Mahler]

No, I'm just really taking it in, Cindy, because, you know, I love listening to M and A people talk. I find it so fascinating. No, I do. I do have a question. I think I've asked Dylan this before with AI right? In this. Obviously AI and probably Pixar's opinion is not going to remove from the creative process, but there is some AI and I'll call it more machine learning that you do kind of use in either post or pre production, like denoising and things like that. Can you talk a little bit about some of those? If you can share some of those processes that you use?

[Dylan Sisson]

I think AI is such a big term. Right. It covers so many things. Right. So there's generative AI, which is one thing. And then we also have kind of these narrow use cases of AI like machine learning denoising, where we started using that on Toy Story 4. And that allowed us to not just render things faster, but render things that we otherwise couldn't render. So it allowed us to make new kinds of images for Toy Story 4, like the antiques Mall and the chandeliers in there. We couldn't have rendered it without the machine learning denoiser. This was technology that was developed at Disney Research. And Disney Research is an arm of Disney, which supports all the studios from W. Das, Walt Disney Animation Studios to ILM to Pixar. And we work really closely with them. We have an internal conference every year called Disc Graph where we share what we're all developing at the different studios. Disney Research takes a lot of the hard problems that need thinking about. And for us to make these movies and they thought about denoising a bunch and came up with a solution that is temporally stable. So it animates with low frequency without any artifacts and doesn't have low frequency artifacts, which are kind of problematic for denoising. And it allows us to render instead of full convergence like 4000 pixels per sample or 4000 samples per pixel to about 60 samples per pixel at Pixar. And with that, that saves us a tremendous amount of time. We had a previous compute based denoiser before that, which helped us out. So we could render maybe 5001000 samples per pixel. But this really, let me put it this way. There's people at Pixar who've worked in production a long time. They're salty, they're not impressed by anything. They've been in the trenches and I've heard them call the Denoiser magic just transforms what they can do and when they can go home.

[Cindy Oli]

So Dylan, how does XPU know, like what work goes to the CPU versus the gpu? Like how is that determined?

[Dylan Sisson]

It's sort of abstracted. So as far as the user goes, you can, you can basically pick which one. If you enable, both the jobs are going to get allocated based on, you know, sometimes the type, the type of operation and just kind of sent across trying to get the best type of operation for the CPU and the best operation for the gpu. But it's on an abstraction level from XPU standpoint.

[Cindy Oli]

So how do you guys test the software across the different workstations? Right. So I know you're doing work on some of ours and you've got some other brands deployed, like how does that testing work?

[Dylan Sisson]

It's critical for something like xpu. Right. And the more environments and the more hardware we can try it out on, especially with the GPUs, all the better. I think for Toy Story 5, when we were rolling out XPU on production scenes during production, like that's basically the best beta tester that we can have, Right. So we were discovering all sorts of little nuances between the images that Res was delivering and the images that XPU was delivering. Sometimes XPU was making a better decision and actually tessellating geometry in a better way. And Res might not have been. There is situations where XPUs doing texture lookups in a more efficient way. So we get more detail when we, because we're using a MIPMAP system, so we get a higher resolution texture and more detail in the textures. There's also things where maybe XP wasn't outputting the exactly Correct. Right. Secondary pass for denoising. So that's something we're like, oh, well, we'll fix that. So there's a lot of ongoing testing on that level. There's an ongoing testing on. We have new GPUs on the farm, so we can have a GPU based render farm. It's very kind of tiny and we're rendering on that and testing that to see how that works and discovering things that don't work. And then we. And then we fix them, but. And then we have our regression tests and all that good stuff too.

[Cindy Oli]

So how does your team feel about, you know, local GPU compute versus cloud compute for rendering? That's something that comes up a lot, especially in this industry.

[Dylan Sisson]

Pixar is interesting because if you go around to the TDs and into their offices, we have a lot of people that in production that are running Teradicis. So we have these VMs that just allow an artist to check out their server for the day and when they log off of it, it turns into a render node at night and they get allocated differently for different artists, for different TDs, and at night they flip over to becoming part of the render farm. Sometimes we do some cloud bursting for production as well, and that's definitely helped out. It's, you know, Blinn's law says it's a corollary to Moore's Law. You know, Blinn says feature film frames will always take the same amount of time to render, no matter what, because we keep asking the computer to do more and more and more and the director wants to see more. It turns out that it's actually taking longer now to render frames for Toy Story 5 than it did for the original Toy Story. Although if we rendered Toy Story today, we could probably do it at 64 frames a second. It's an interesting relationship. And I think also what you're getting at is this idea of, you know, kind of the local artist is working on a shot interactively and then the lighting's done, the shading's done, and then we want to send that off to a farm render that. That's true if you're working at Pixar, or if you're just an individual freelancer or even a student, and being able to harness all the hardware that you have on your local workstation. If you have a GPU and a cpu, why not use them all on focusing on getting those pixels as fast as possible, then maybe you want to render in the cloud. You can render the cloud, you can choose whether you want to render on the GPU or CPU and kind of pick whatever fits your needs at that moment.

[Cindy Oli]

So I'm going to pivot for a second to the fun stuff. Let's talk about the way that you guys give back. I want to start with the art and science fair, because that is one of my favorite traditions at Siggraph. And for people who haven't been to Siggraph, do you want to kind of give everyone an overview of what that is? You know, how does it help you serve the Renderman community and just kind of give us some details on that?

[Dylan Sisson]

Yeah. The Renderman art and science fair is something we do during the week of Siggraph in the evening, and we try and get a nice venue, hopefully kind of a rock and roll ven. We allow people to come and socialize. We have our signature drinks, including the render Manhattan, and we let everybody hang out. We have demo stations so people can walk around and see what we're working on. Then we have kind of a main stage show where we reveal our biggest messaging for the year. And at the end, everybody gets a walking teapot. You may have heard of those. You know you have a teapot, right, Logan?

[Logan Mahler]

I do, but we need to talk about the Dell saucer to go with the complimentary teapot.

[Dylan Sisson]

The Dell saucer. Okay.

[Logan Mahler]

No, I mean, I've talked to you about it. You kind of shrugged me off. But, like, now people are watching. You can't back away.

[Dylan Sisson]

A saucer with legs?

[Logan Mahler]

Yeah. No, it doesn't have to walk. Like, it's more stationary, but, like, you know, as a stand for your teapot, it's something Dell and Nvidia inspired you.

[Cindy Oli]

I think you were the mastermind behind the teapots. What inspired you? And they've become, like, a phenomenon. The line for teapots at Siggraph is usually around the building. How many are there? I'm slowly trying to build up my collection. I've priced them out on ebay. It's out of budget.

[Dylan Sisson]

They really took off and got a life of their own. I designed the first teapot in 2003, and I was supposed to create a brochure. And I thought, well, every brochure every year always says faster, more robust, et cetera. And I was like, what if we did, like, a toy? Just a giveaway for the community, because I'd done some collectible toys in the past. And so I started emailing toy manufacturers. I found one that would do a Utah teapot. I was like, can you put a walking mechanism in it. And they're like, yeah. And then they sent me the prototype, and I was like, okay, let's do it. I pitched it to my boss, and he said, is it more expensive than printing a brochure in the States? And I was like, no, it's actually cheaper. So it turned out that my little project took off. It took a couple years for people to actually realize what was going on. But once they did, the teapot became very popular.

[Cindy Oli]

Oh, even the plastic bags, I was told, you know, you've got to get the lowest number on the little tag. You've got to keep them in the plastic bag. And it can't be any plastic bag. It has to be the plastic bag that the teapot came in. And apparently there is a type of bag. And so, yeah, I'm now a teapot community member. I think.

[Dylan Sisson]

I think people love their teapots, and people put them on their desks. I talk to people at Siggraph. We have our teapot museum. And people will come up and they'll say, yeah, this was my first Siggraph back in 2014. And I got this teapot, and it's sort of this marker initiation for people, which is great. But I've also seen them go on ebay for over $600, which is more than a license of Renderman, which I think is a. Which is kind of a funny, funny metric.

[Cindy Oli]

Aside from EMT pods, aside from the art and science fair, you also have the challenge. And I think, you know, I love that because it's really inspiring to see so many artists. It's a global event. Tell us about the challenge.

[Dylan Sisson]

Yeah, the Readyman Challenge was something that we. I think we're on the 13th challenge now coming up, and we've been doing it for a number of years. And the idea was to basically provide some assets, invite anyone in the world to participate, become a challenger and take those assets and tell a visual story in a single image, and using Renderman. And now we've our last challenge. We encourage everybody to use xpu. And the level of entries was really astounding. We had more than 200 entrants this year. And we have some great prizes from our sponsors, like Dell. We have the great hardware people can get if they win. We have. I think it was over $40,000 in prizes for first place last year, which was astounding. We also have mentorship, so we provide mentorship from our team for people that want to render volumes or want to render hair and want to figure that out. We have other teams from like thoughtful 3D and ozone who have provided mentorship for character design, mentorship for like developing ideas. And it's really turned into something that was bigger than just a contest. It's actually been a platform where winners have been able to put it on their resume and attribute them getting jobs to having participated and become a finalist. And a challenge which was really cool, including two people at Pixar, which, you know, it's nice to have a part of something that's helping the community too. So it's been a very, very cool thing and we have a lot of support behind it, especially from Dal as well. So thank you.

[Cindy Oli]

What advice would you give to first time challengers looking to up their creative and technical skills?

[Dylan Sisson]

You know, the border of the barrier of entry is fairly low. Like if you have a computer, you can get Blender, you can use Renderman non commercial and you can try, try out XPU and see how it works with the assets that we provide for the challenge. The next challenge is going to be starting at Siggraph this year. So that's in July. And if people want to join, you know, I encourage them to jump on our discord, participate, ask questions. The community is very good about helping people kind of develop their ideas. It's just a really fun way to explore storytelling in a single image with some of the latest cutting edge hardware. That's Renderman XPU that we have in non commercial Renderman, same XPU that we're running at the studio. So it's cutting edge stuff that people get to play with.

[Cindy Oli]

This has been great, Logan.

[Logan Mahler]

So how we ended up doing what I like to end with, because I know you need to get going, is basically take 30 seconds, tell everyone where they can kind of find you, you know, where they can check out xp, where they can download the non commercial version and then we'll, we'll call it a day.

[Dylan Sisson]

If people want to follow up, say hi, I'm on LinkedIn, I'm on Instagram, Real Dylan Sisson. And if people want to try out xpu, you can download it today. You can use it with any of the DCCs that we support, that includes Blender, Maya Katana and Houdini. And it turns out we have a website, it's called renderman.pixar.com and you can go there and download non commercial Renderman and it does what it says on the tin. You can use it for non commercial products, projects and it is free of watermarks and limitations. So you can try that out.

[Logan Mahler]

That'd be a great project for Cindy to actually learn some software in the industry she supports. That'd be a good project for her. That's. Enter her in the challenge.

[Dylan Sisson]

I'm not getting in the middle of this.

[Logan Mahler]

I know. You don't. You don't. You don't want to. You know what to. So, with that, we're wrapping up reshaping workflows with Delpro Precision RTX Pro GPUs. Dylan, it's been a pleasure having you. Can't wait to see you at GTC and seagraph. And that's it until next time. See y' all later.

[Dylan Sisson]

This podcast was produced in partnership with Amaze Media Labs.