C (2:40)

Amazing. Cindy, dare I even ask you to introduce yourself?

D (2:44)

Hi, Cindy Olivo. I'm the marketing. Marketing manager for media and entertainment. That includes game development. I actually am a gamer. I'm also trying to get Logan's daughter to be a gamer as well. I think I'm successful. I've been successful at that.

C (2:58)

If you ruined her D1 softball scholarship versus gaming, I'm going to be very upset.

D (3:03)

I gave her a video game that Logan was unable to figure out how to install on her Switch. I've been playing Mario all weekend.

C (3:12)

Yeah, exactly. Refusing to go do our workout. Anyways, doesn't matter. Thank you, Cindy. You're not a good force in her life. So, game development. So obviously people play a lot of video games, whether it's PC, console, etc. I'll throw it to Paul and then. Or Rick, whichever one of you want to take it, you know. So we'll get to kind of the Nvidia side, but from an ISV kind of standpoint in software and just game dev. Is this unreal? Is this Epic? I kind of describe the, the general kind of workflow ISV for those out there that don't understand what goes into game dev. What's kind of the base level foundation to create a game.

A (3:48)

Please, Rick, feel free to jump in and correct me when I, when I, if I, if I step off the path at all. But so there's a few really key things that go into creating a game. So you generally have a game engine, so something that will power the actual game logic, bring together all the models, the art, and kind of be the, like the engine driving everything that's happening in the game. And so like you just shouted out unity, Unreal. Maybe Godot. There's a lot of popular engines out there, but the, the two that do sort of dominate the market are Unreal Engine, which we're now on the, the fifth Unreal Engine that's produced by Epic Games, and then Unity, which is produced by Unity. And so the engine is kind of like one of the core components, but there's all this. There's an entire ecosystem of ISVs that operate outside of the engine, but that. Where you might design assets or you might be writing or designing actual quests before you make it all the way into the engine itself. And so examples of those might be things like Adobe Substance Painter or, or Maya from Autodesk or potentially Blender to do some sort of modeling. And then you Also have an entire set of ISV providers that exist on the service side of the industry. So for example, if you are Riot and you. Well, I don't think Riot uses them, but if you are an online live service game, you might want an anti toxicity service that analyzes the communications on your game and, and jumps in and says hey, you can't say that you're banned, that's not an okay username. And so that's something would be like, something like talks mod. And the way that I think about it is we have these kind of pre engine ISVs, things that you use to build your assets. We have the engines themselves and then of course we have the ISVs that you might use to deliver the services or the live service ISVs and I'll toss it over to Rick to pick up any broken pieces that I want.

B (5:52)

But yeah, I mean if you look at it like the, the asset generation is, is a very heavy part of the build. You know, creating all the characters, creating all the environments, creating all the, you know, from the concept art through the models, the textures and the animation and then combining them all within the game. And so within the game engine, you know, that's kind of where we bring everything together. And then as was Paul was saying, you know, we've got the, the actual live service for some games, you know, there's a whole bunch of microservices that are being run. In a lot of cases these large games, those services are stood up by the individual companies. They design their entire, you know, service back end and we work with them and different cloud service providers or their own private data center distribution in order to make sure that those services are up and running, whether it's toxicity monitoring or it could be live translation, a whole lot of things that we can do.

C (6:40)

So that makes sense. So when you're developing game obviously you have kind of probably a bad term for it, but the game engine, the orchestration engine, there's different services providers, whether it's unreal unity, all of this. So this I'll start with Rick and then I can go to Paul is that obviously Nvidia accelerates most of those softwares, at least from my understanding. But like from the software SDK standpoint, where is Nvidia involved in kind of the game development side? You mentioned demo before, you mentioned translation where kind of the key outside of the gpu. We'll talk about that in a second. But from a software standpoint, where does Nvidia fit in in the game kind of development kind of workflow yeah, there's a few parts.

B (7:19)

If we're Talking about the ISVs that are utilized for say, asset development, there's a lot of our SDKs that are pretty typical. You know, whether you're walking with an Adobe or a Blender or an Autodesk, they use a lot of our work in order to get graphics acceleration within their applications. So at a low level. But then there's, there's a lot of work that's being done at the large major studios. A lot of them will have their own game engines. So if we're talking to some of the really, you know, long running game franchises, they've got their own engines and they'll use those same SDKs in order to get, you know, RTX ray tracing or to use DLSS to accelerate the game at runtime. And so we've got a whole team that focuses on game technology that will be put into those game engines. The focus of the team that Paul and I, you know, are looking after are more how can we help the developers and the, and the, the artists that are working actually on the game development rather than accelerating the game engine. And a lot of that is okay. The use of AI is kind of a hot button topic, you know, around game development. And so we're helping a lot of the studios look at how can they use AI responsibly in order to make a lot of the game processes much more efficient. And a lot of that could be in customizing large language models to help with content understanding. So we've got an entire stack that we utilize across industries to help with model AI training and inference. And we bring those to game developers and we want to kind of have them, you know, kind of treated on that, that same playing field. You know, these, these tools that are available to, you know, a large service now or an amdocs is also available to a game studio.

C (8:53)

Fantastic. Paul, anything to add?

A (8:56)

Yeah, I mean, I think Rick pretty much covered it, but there, there, I, I do really want to emphasize that like for us specifically, we, we think of that divide as oftentimes coming when runtime for a game starts, that's, that's when we pass out of our domain. And so we are really thinking about everything that we can do to make the lives of the designers, the artists, the pe, the, the QA professionals, the infrastructure IT professionals, the programmers, the, the, the decision makers in the, the gaming industry easier. How do we make it so that they can move faster? How do we make their, the, the timed iteration quicker? How do we use Nvidia technology and of course running on Dell machines to really make this industry as a whole faster, accelerate it. And I, I personally am so excited to be in this industry at this time because if you look at things like what, what just happened at Unreal Festival earlier this year where Epic got on stage and Tim Sweeney said we're pushing the AI button and we're, we're finally introducing like some, some studios are, are finally being willing to introduce some of these tools into their workflows and into the games themselves to actually make the industry faster and, and, and help. Help people do their jobs at, at a higher level.

C (10:11)

That's very cool. Let me ask, you brought up AI. I was going to save it, but you know, this isn't. It's. It's Nvidia, it's AI. We got to go and touch it. So give me. And whether we start with Rick or Paul, whatever works, give a couple of use cases like you just mentioned, you know, Unreal Fest, turning the AI button on, what are a couple of those workflows? But then also give the ultimate end result that someone playing the game would see. And I'll give you one that I know is non player characters, right? Being able to create kind of a rag, you know, database where they can reference. So they say a little bit different stuff, but it's within the context of their character. Hopefully that's not the only one. If I just stole it. I'm sorry, but what are some of those, couple of those AI ones. So customers that are people that are actually playing games, like Cindy's husband who plays a lot of games could be like, oh man, that is driven by AI. I had no idea. That's super cool.

A (11:00)

I'll start off.

B (11:01)

I think if we think about AI, like there's a lot of people that are looking at it from. There's a lot of AI that's novelty, you know, like I can use an image generator and I can get a teddy bear riding a unicorn, but I can't control that image. So we look more and there are,

A (11:14)

there are people that are working on

B (11:15)

that and bringing some of those technologies more to a usability, you know, point for production. But when we look at like workflows of say a lot of developers working with a custom code base and you bring in a new developer, great developer, not familiar with your engine. If you stood up a rag that is familiar with the engine, you can find functions and things within your code set that you know would be easier for someone to find. Being able to reference their entire slack Channels, email, threads, in order to get, you know, the undocumented documentation that might be around. So it makes the lives of a developer a lot easier. But also as you go into things, you know, like non player characters, where it's really exciting is rather than spending a lot of time in some cases just writing dialogue trees, you know, the narrative storytellers can actually develop really complex backstories for these characters. And so depending on the interaction, you know, they get to glean on this entire history which in, in some portions of a game, you know, for a character who may only have a few interactions, it might not have been that exciting, but that character can now have a much, you know, broader experience, you know, when being in an interaction because they have the ability to do more than two or three lines of dialogue. So it can make a rich and unique experience for people.

D (12:28)

I have a question, Rick. So in the previous generations of GPUs, I know that there was architecture that was built into those specific specifically for game development, especially around ray tracing and path tracing and things like that. What's new on the new Blackwell cards that is specifically for this industry?

B (12:47)

A whole bunch of stuff. So I guess the great thing is, is when we look at the professional cards and the consumer cards for gamers, it's the same chip. So all the technology that we have, plus some. So like we have, you know, error correcting memory on the enterprise boards, which is really big deal for us when we're doing large simulations and you have error that you know, collects over time. But some of the stuff that we do, or you know, enabling the new DLSS for, you know, being able to do multi frame generation, There's a lot of interop between different fixed function hardware that allow us to do things, you know, like optical flow acceleration. We're taking advantage of the tensor cores even more than we did before. We're seeing large appetites for memory on the consumer side, which means double that for a developer. We can get into some good use cases with that. But I think the large thing is we also see the generational leap of just overall performance. So I remember in the old days, you know, you'd get like 5 to 10% performance increases for generation. Generation. Like yeah, yes, it's like 10% more proficient these days when we're getting 15, 20 to double the performance. Generation, generation, but it depends on the operation that you're doing. So like some rasterization graphics, you know, you might, we might see a 15 to 20 to 30% performance boost but then some ray tracing operations. Because of the utilization of the RT cores, we might see a 2-3x performance boost generation to generation. And DLSS gets better and better with each one. So we've switched over to some diffusion based models. The, the results are fantastic and especially when we add the multi frame gen. So you know, we look at it, you know, we're like one of every 16 pixels is generated by AI. And where the advantage is, especially if you got these really dense scenes and the resources required to render at a certain resolution are really high. To double that resolution or triple that resolution requires such an amount of resource that your frame rate would be just through the floor. But now being able to inference that the inference load doesn't scale the way that a rendering load does. So the inference load for the resolution of the image is mainly based on the resolution of the image, not the complexity of all the assets and stuff within the scene. So we allow developers to have much more complex and engaging environments, but still have them run beautifully in high frame rates.

D (15:09)

I really like that. I think that is how we differentiate as well on our side between our consumer products and then our enterprise level products. We take that kind of made on played on approach. If you're playing the game, you want to be on the Alienware. If you're making the game, we want to make sure you're on pro graphics on Dell Pro Max. And so it seems that you guys are taking that approach. And there is real differentiation between the consumer level cards and the enterprise cards.

B (15:35)

The systems are completely built differently, different thermals. And you're going to be hard pressed to find a gamer who's going to need 96 gigabytes of memory. But we have developers who like that's, that, that's hitting like a good sweet spot for them. And this is a holdover and holdover, but it carries over from other parts of the media entertainment business where you know, we were seeing the kind of like the gigabyte requirement for most renders for feature film level work is around 60 to 70 gigabytes. And you know, beyond that you hit points where, well just the cost of like the asset load, the network and everything, you know, kind of gets untenable. So you don't want to go beyond that unless you've got like this one

A (16:14)

scene where you're destroying New York with this giant wave.

B (16:17)

And okay, yeah, you might have 250 gigabytes of data, but in most cases you want to kind of keep it within that budget. So now we have a single GPU that can fit these really large scenes into a single frame. So that can be used for these in these great, you know, rendering situations. But when we go to gaming, the issue that we saw a lot of times is people are maxing out their GPUs, loading the game and they're playing, they're loading the game, that's great and they can play it. But now I gotta go back and do, do the work and so I gotta reload all my assets again. And so just that, that iterative process of I've loaded, you know, I've loaded Unreal, I've loaded all my assets and my assets are not optimized at this point because I'm still in development. So that's taking up like 30 gigs of space. And then I wanna play test it. If you wanna play test a build in the old days, well, you shut everything down, you're on an offline build and then you play it and then you spend 15 minutes loading everything back up again. With a 96 gig frame buffer, we can hit a button into builds and you play it and then you close it when you're done and you still have your environment open. So it's great time savings just from a developer point of view. And this is for level designers, this is for gameplay programmers. People who are running large scale simulations or just running multiple applications, like a lot of artists these days are, are so versatile that, you know, they're jumping back and forth between, you know, I'll, I'll start something off in Maya and maybe I'll take it into, into zbrush and I'll touch it up and then I'm going to take it into Substance Painter and I'm going to work on it. And if I'm adjusting the UVs for something, I got to go back into this application. So closing things down doesn't make sense. But when you've got the memory in the system and you've got a GPU that can handle it, you can have all these applications open and you can still see the final result in the Unreal engine all in real time.

A (17:57)

The other thing I'd call out there is like the huge benefit of having such a large amount of VRAM is that we can hold larger models in memory. And so like if we're running, running around between all these different apps, like, of course you can also have Comfy UI open and you can be able to, you know, do diffusion models, then bring them directly into Adobe Substance Painter. But Logan, you originally asked about things like rag and smart NPCs and I think One thing to think about is, you know, we're getting to a point with LLMs where we have, you know, hundreds of billions of parameters and your standard consumer card just can't fit that into, into memory. And so if, if we're looking at something like more like a, a more reasonable like 70 billion parameter model like we had, you know, even just a year ago, these are things that you can really reasonably use to power rag pipelines on your device while you're making quests or while you're writing code to do code assist or potentially even while you're looking up documentation. We have a great use case of loading in the entire Unreal Engine documentation into a RAG pipeline and using, using it to help developers as they're actually building in Unreal. But the problem is how do you take those LMS that are, you know, they're, they're the size of the, of the model itself is, will exceed the buffer for these consumer consumer cards and, and get them into the game itself. That's when we start thinking about things like model distillation where you know, you can take the weights of the model and go in and look at what are really the important weights here that, that add up to something that can have, you can have a conversation with that doesn't really. We don't need our NPC characters in games to know about Motorola Flip phone if they're in the middle of a medieval world. But you can go in and specifically just pick out the parts of the weights that you need in order to let them know about maybe a dragon or a unicorn or, or a long sword. So we've, we've seen this ad Nvidia where we actually worked with a company in Zoy to create Smartzois. So this distillation produces not a large language model, but a small language model. And those small language models are small enough to actually run on the consumer product. But in order to get to that point, you definitely need to be working with enterprise hardware. And so this is kind of the, the aspect of it where it's exactly what Rick said, you know, with the growing, you know, sort of sort of memory buffer of, of all these consumer products, of course the enterprise products have to be that much bigger in order to be actually create these experiences.

C (20:24)

That's a really interesting point. Like what is. So you kind of brought up, you know, we've talked Blackwell 6000. I mean obviously the workflow and kind of the, the device of choice is obviously changing, right? I mean we were ADA, now we're Blackwell. I mean, I'd be curious about GB300, but we'll get into. That's a whole other question. But what is the typical. Because I've seen Anemone, Cindy drags me along to a lot of shows and a lot of it is, hey, we like, you know, rack mounted workstations. We like to kind of piece and part out potentially the gpu. Like what is kind of the, the industry standard for, you know, a game developer? Like what? Yeah, is there an industry standard? I don't know. Like what or what would you recommend as someone who's watching as a game dev, what would you recommend to them? What would be the configuration, GPU wise?

B (21:13)

I think the, the big advantage that people have with the space right now is that we understand that it's not one size fits all. So that just like we have multiple consumer, you know, consumer cards, we have multiple enterprise cards, then we also have different server types depending on whether you need the heavy iron for model training. And this is just going to do AI inference all day, all day long, you know, long. Versus I need graphics performance. And depending on your size and your IT infrastructure, you might want to have something that, you know is smaller and more manageable, where people kind of like, I'll say self managing and they've got, you know, the carpet cluster and everyone's got a workstation on their desk or under their desk. Versus we want to go to a more of a managed deployment solution and we're going to either rack mount everything or we're going to move people onto servers. We're seeing as studios increase in size, there's a lot of appetite for virtualization. With virtualization you can a lot of cases save money versus buying a whole bunch of large workstations. And in a lot of cases, not everyone's going to take advantage of all the power of what you're putting into that workstation. But if you build, spend a little bit more money on a server platform that you can put multiple users on 2, 4, 8, 16 users, depending on their workloads, you could dynamically allocate the resources that each one gets. In some cases you'll have, you know, animators or modelers who are primarily working in something, you know, like Maya or Blender, and their GPU resource requirements might be very low. So while it's great for them to have a 6000 GPU, they're not going to be utilizing it, or at least not all the time. So being able to put a card like that in and split up between 2, 4 users is kind of a game changer for studios where they can build out an infrastructure that, you know, during the day I've got four people that are working on this server and then at night the memory footprint is so high that I can use this for model fine tuning. So it really does make it to be kind of like that universal server when we're looking at the RTX Pro line right now.

C (23:16)

Makes total sense, Paul.

A (23:18)

Yeah, so the other thing I would, I would definitely think about is, is there are, there's definitely right sizing for different sizes of studio, different types of game. I mean this is something that Rick and I were just talking about the other day about the, this. If you're, if you're dealing with a massive, you know, 20 kilometer scene with, with really intense levels of detail, obviously you, you're, you're probably going to need the big iron just to render things local, right? You're going to need a really powerful workstation with a, with a powerful processor and an RTX Pro 6000 in it. But if you're dealing with, you know, smaller scenes and you're, and you're dealing with, with less like, I guess I would say dynamism in there, like you, you could probably fit a 5,000 into that and take advantage of, you know, less than, you know, you don't need the full 96 gigs of VRAM. But the other thing I'll call out is, is I think that this, let's think about how fast the gaming industry is changing. They're just in general the, the complexity of games in the past 10 years alone has, has. It feels like it's almost double, maybe even tripled. Right. And so we're also looking at expanding development budgets that at times aren't keeping up with the actual revenues of games and the actual work that's going to go into creating what is the what. What the market actually wants, which is more and more complex games with, with a lot more work put into them on tighter budgets and tighter timelines is going to require hardware that, that is not just ready for what's going on now, but is future proof. And so I think that's just one thing that any company that is, you know, currently in the process of evaluating hardware for their game developers, for their artists, for their QA testers should be thinking about is, is there will be AI native workflows that happen. I just saw a, something that looked like, you know, the equivalent of like a cursor but for 3D, 3D modeling and scene composition and something like that will have you know, a very, very intensive memory requirement. And so I think that just in general, like we, we can right size for today, but we also need to be thinking about how quickly this industry is changing and, and how a lot of the tools that are currently being used are going to have, you know, much more AI intensive components in the next three to four years.

D (25:34)

A while on that, where do you see the industry going? You know, games are becoming more complex. We're seeing a lot more integration, AI integration into the development process. What does the future look like for gaming?

B (25:48)

I think it's going to hit a rebound in short order. And I am not an expert on the gaming business. I'm more looking at the technology side. But I think what's happening is people are going to start taking advantage of the tools and the processes that are just now becoming available kind of at a production scale and that's going to help drive efficiency. The quality of the games is going to keep getting bigger and the requirements for the, for the people within those games is not going to get any smaller. So we're going to have, you know, I don't see production budgets necessarily shrinking, but I don't see them growing as exponentially as they have been. I think the people that are working on these games, each one is going to be much more efficient and much more effective at their primary job. Like some of the stuff we were talking about earlier, like when a developer is wasting 10 minutes out of every test cycle, they're wasting an hour to two hours a day, you know, and if we can make that much more efficient. So you know, you're spending, you know, you're getting, you know, 20%, 25% more efficiency out of every single developer. Your costs are going to go down, which means your, your, your production is going to come in. So we don't see an overall contraction, I think, in, you know, the size of the teams. I just think we're going to see the game schedules being able to be brought in because they're more efficient. But then also the deployment of those games, being able to target the right audiences, being able to, you know, deeper engaging interactions within the games, I think are going to start increasing. And as all that does, I think it's just going to, I mean, games are currently, you know, the single largest revenue source of entertainment out there. You know, it eclipses what we do for, you know, film and TV production. So I don't see that slowing down. I just think, yeah, as cost costs have been skyrocketing, it kind of needs to be reined in and we think it's going to be, you know, we're looking at it like, how can we help the ecosystem gain efficiencies? And so we're, we're kind of at the low level of everything, right? You know, like we don't make any of the software that the creative community uses. We don't make the end software that the developers use. We want to accelerate it and make it much more efficient. So, and we're also open to ideas. So, you know, like, the more that we can find out from people of what's missing in the ecosystem, you know, what can we accelerate, what can we help with adoption, you know, what are new technologies that might, you know, leapfrog, you know, current, you know, tasks that people are doing. Like, we're, we're open for that research.

A (28:11)

The other thing I think that we, we talked about ISVs earlier, but we didn't talk about cloud service providers. And I think one thing that's really important to think about for the actual runtime games and the way that the industry works right now is that so many of these games are live service games. You release, if you release a game a year, that game still has, you know, a five plus year lifespan where you're having to maintain servers, where you're having to release new updates and code. And I think that when we, when we see things like inference starting to show up in the games themselves, like for example, like to go back to Epic where, where they put Darth Vader into the game, I think just for a few players, for about a week, you start to see that there's going to be, you know, a massive overhaul in the infrastructure that's needed to operate some of these live service games. And obviously, you know, Dell and Nvidia hardware are going to be required to both support that infrastructure, but develop all of the code that's going to run on that infrastructure. And that starts with things like, for example, you know, we have an upcoming product like the GGX Spark or the GB10 from the Dell side that are going to be necessary for people who are just working on their desktops to be able to actually do levels of AI development and fine tuning and to work with these, these tools before they actually even make it into the cloud. So, so I think just when we think about how the industry as a whole is going to change, I think we're looking at, at least from my perspective, a lot more inference ending up in the game runtime itself. And in order to support that, a lot more sophistication and people on the development side who really understand how this stuff works and who are equipped to make the changes before they even end up in the cloud.

B (29:55)

I think to add to that, there's also being able to go back with these live service games and revisit what may have been seen as a solved problem and make it much more scalable, Bring the efficiency up, bringing the costs down. We did a big thing with Riot Games or their toxicity monitoring system. We were able to get them a 70% savings and drop their latency down to 1% of what it was previously running, basically the same code. So we optimized it and they're still being deployed, you know, you know, through cloud service providers. So everyone is still basically happy. But the idea is, you know, we can look at some of these microservices that are going in and our job generally isn't like we don't do people's research. We want to help you scale and go to production. And so a lot of times, you know, like it's, it's easier to stand something up and test something, especially something that might be a low resource job. You know, that you can run on a CPU and do simple inference and this is going to be great. And I have like get 10 people per, per CPU on this. If we can transform that and run that same model and now get 500 people per GPU on that. The GPU is more expensive, but the overall cost really comes way down on per concurrent user. So we like to look at workflows like that. So where can we find those, those operational savings so that the game is exactly the same, the experience is overall better and the, the cost of the overall studios goes way down.

D (31:18)

As we wrap, I have to know what's everyone playing?

B (31:22)

Paul?

A (31:24)

I mean I, I'll be honest, I. I am a really big indie gamer. I just finished playing Dredge, which is a, a fun little fishing game that has like a Cthulhu horror mythos embedded in it. But I'm, I'm also, I, I have never played red Dead Redemption 2, so I'm just starting to play that as well.

C (31:42)

I figured your favorite game would be Attack of the Killer Pancakes.

A (31:48)

I've never.

C (31:48)

Inside joke. Inside joke.

D (31:50)

I'm wrapping up Indiana Jones in the great circle. What about you, Rick? Playing anything?

B (31:55)

I'm doing a few things with, with my kids. So we, we got the switch too. So we're playing some Mario, but also bouncing between Path of Titans because dinosaurs are popular around here and Subnautica. So we're hoping to finish Subnautica and looking forward to Subnautica too.

A (32:09)

So that's um.

B (32:10)

And the. The old stand was Roblox is still hugely impactful around here. So it's another day goes by where someone isn't jumping on for. For a little bit.

D (32:19)

And Logan, nothing on your end?

A (32:22)

Nope.

C (32:23)

No, we'll work on you. I just, I. I don't have that. Eye and hand coordination is a problem. I think it's probably because I'm blind as a bat, but that's neither here or there. But. So, yeah, fellas, this was awesome. Really appreciate the time. So we'll start with Rick, then we'll go to Paul. What I like to do to recap the episode. Take 30 seconds each. Pretend someone has just joined and you need to give them the, you know, the elevator pitch on what they need to know about kind of gaming with Nvidia. We'll start with Rick. 30 seconds.

B (32:55)

Go Nvidia. On the Enterprise solution side, we are here to help people take their research and their development and put it into production. You know, whether it is, you know, new tools for content creation or releasing a microservice in the cloud for help for live service game production. You know, we want to take their development pipeline, make it much more efficient, make it much more inclusive and get everyone the tools that they need to get the. Get the game done on a budget.

C (33:19)

That's amazing, Paul. Can you beat that?

A (33:21)

What I would say is that we're powering the full game dev pipeline from artists to designers to programmers to qa with Dell and Nvidia solutions together, they're going to have everything they need to actually complete the game on their desktop. And that means no switching between apps, that means no closing things so that something else can compile or you can start a play test. And with the game industry changing as fast as it is, every single bit of time matters. So with these two pieces of hardware together, we're equipping the game industry to be ready for any future change and to meet it head on.

C (33:56)

It's fantastic. This is great. Well, Paul, Rick, really appreciate the time. Where. And I'll go to Paul where if they want to learn more about in Nvidia gaming and the underlying kind of models and things that you provide, as well as the GPUs. Where can they go on Nvidia? And we can link down into the description of the episode, but where should they go to learn more?

A (34:15)

Yeah, so we have an enterprise game development Nvidia industry page. I'll go ahead and share the link to with Logan before the episode wraps. And then we also have build.Nvidia.com where all of our NIMs and blueprints and just various different AI models live that developers can pull down and play with. Or they can also just play with the APIs themselves to experiment with all this stuff before they commit to it.

C (34:41)

It's amazing. Well, another episode, another new subject touched. Really appreciate the time, Paul and Rick, and all of your insight into kind of the enterprise gaming space with Nvidia. So with that, this is reshaping workflows with Dell Pretty Pro Max and Nvidia RTX GPU signing off. Until next time, keep your workflows running locally and we'll see you on the next one.

B (35:21)

Do what you want. This podcast was produced in partnership with Amaze Media Labs.