Clips
News
More from NVIDIA GTC 2025

Tony Paikeday

Senior Director of AI Systems

NVIDIA

Nvidia partners with Equinix to deliver ready-made AI factories

Nvidia Corp. hopes to continue its dominion over artificial intelligence infrastructure with its latest generation of DGX SuperPOD.The release of DGX GB300 SuperPOD comes alongside the fruits of Nvidia’s ongoing partnership with Equinix Inc. The digital infrastructure company will manage Nvidia’s Instant AI Factory, a service for building AI-ready data centers in a relatively short time frame.Nvidia’s Tony Paikeday (left) talks about Equinix’s involvement with Instant AI Factory

play_circle_outline New DGX SuperPOD with GB300 for AI reasoning and deep learning models

play_circle_outline Equinix instant AI factory for easy deployment of infrastructure solutions

Info
Transcript

Tony Paikeday, Nvidia

Tony Paikeday

Senior Director of AI Systems NVIDIA

Exploring AI Potential with Nvidia's Tony Paikeday at GTC 2025

In this insightful segment, Tony Paikeday, a distinguished leader in artificial intelligence solutions at Nvidia, joins us during the Nvidia GTC '25 event. Hosted by theCUBE Research's Savannah Peterson, we explore Nvidia's groundbreaking advancements in AI infrastructure in collaboration with Equinix.

Paikeday, with extensive experience in AI computing, discusses the evolution of AI from its early stage to its current state of dynamic reasoning and intelligence. They highlight the s... Read more

explore Keep Exploring

What role does Equinix play in helping businesses deploy the DGX SuperPOD? add

What partnership has been formed and what benefits does it provide in terms of deployment and operation at scale? add

bolt Powered by CUBE AI

Tony Paikeday, Nvidia

Savannah Peterson

>> Good afternoon data community, and welcome back to beautiful San Jose, California. We're here at NVIDIA GTC with our exclusive coverage from the Equinix Booth. Very excited for our next segment here. Tony from NVIDIA, thank you so much for taking the time.

>> Thanks for having me. Great to be here.

Savannah Peterson

>> Kind of a busy week for you in particular, just the buzz in this room alone. Does this feel like a party where you get to celebrate with all your friends?

>> Welcome to GTC. I mean, everyone I've done like this has been borderline insane, but also incredibly exciting. There's so much going on. We're on the heels of the keynote and we truck out our best, most important news for the world of AI, and this year was no exception. That's for sure.

Savannah Peterson

>> I'm really glad we get to talk about a lot of that. We're here in the Equinix Booth. You guys have been partners for a while. Talk to me a bit about that partnership and the value it brings to NVIDIA.

>> Yeah. You think about AI and where we were, say a decade ago. In the earliest days of building AI infrastructure, not a lot of folks had competency expertise in how to implement the delicate balance of compute, storage, network, power, facilities, security.

Savannah Peterson

>> It's a complex situation.

>> Totally. The partnership with Equinix is a match made in heaven, because they bring the operational discipline, the expertise, the know-how about building next-gen data centers with our compute architecture and our platforms. That's why it's a real one plus one equals three for any business that wants to build scale and operate AI.

Savannah Peterson

>> Yeah, absolutely. I can see why this has been such a harmonious partnership for a while. Jensen just spent two hours on stage, two days, because it feels like he was on stage, spent two hours on stage breaking down all of the GTC announcements. Give us some of your key takeaways.

>> This year started out very exciting with, we all felt like the deep sick moment, right?

Savannah Peterson

>> Yeah.

>> And how that turned the whole world upside down around what you could do with highly efficiently developed models. But the reality, as you heard from Jensen, is that these new state-of-the-art models drive an incredibly deeper, more extended amount of computing because they're reasoning. Multi-layered thinking, reasoning, step by step by step thinking options, alternatives, then creating an answer and presenting that to the user. What the world is realizing and what they heard from Jensen today is that in the age of AI reasoning, you need purpose-built compute infrastructure that has an immense amount of computing power to power those models.

Savannah Peterson

>> Yeah, absolutely. There's a lot of dynamism across the different industry needs within that. Very specifically, literally. Just a few moments ago, big announcement, DGX SuperPOD with GB. Tell me all about it.

>> Yeah. The latest generation of DGX product is the GB300, the DGX GB300 SuperPOD. We've delivered DGX solutions to the marketplace for almost 10 years, nine years at least. This is the latest iteration, but it's built for this age of AI reasoning using the Grace Blackwell Architecture. What's so important is that as businesses think about what is the essential platform, the essential infrastructure to tackle state-of-the-art models that can do deep reasoning, long thinking with incredibly immense memory footprint, with ultra high-speed, low-latency networking, interconnected with storage in a turnkey solution, they're going to look to this generation of SuperPOD, right?

Savannah Peterson

>> Yeah.

>> Equinix is such a key component of that platform because you need liquid cooling, you need next-generation data center facilities. You need ultra high-speed interconnects to the world outside to bring that data in. And so Equinix is proving to be an incredibly valuable partner to help the world's enterprises deploy the DGX SuperPOD with GB300 in their data center.

Savannah Peterson

>> They're the first ones to be able to offer this, I believe. Correct?

>> Yeah. We've been working with them for a while, with Equinix. We've in fact been deploying early production engineering units there. It gives us a lot of operational experience about deploying and operating the system at scale. The beauty of this is part of that announcement is we also announced Equinix instant AI factory. You think about organizations around the world that need access to this system, but they don't want to struggle for months or potentially years to find a home to deploy it, to scale it, to have it managed and operated. Our partnership with Equinix will have us in over 45 markets around the world with pre-configured facilities that are ready based on customer demand to scale up and operate and manage these infrastructure solutions.

Savannah Peterson

>> That's huge.

>> Absolutely. It's going to really shorten the timeframe from knowing that you need this mission-critical infrastructure to actually putting it in your hands and giving you the keys so that you can start and point your data scientists at it. It's going to make it. It's going to have a huge democratizing effect on the world's businesses that want access to DGX SuperPOD with GB300.

Savannah Peterson

>> I cannot wait to see what your community does with this collectively. Curious, just because we're sitting here for a second, you get to see and hear some of the most interesting and exciting real time and also potential future applications for AI. What has you personally most excited for our future?

>> There's a couple of things. You probably heard some of it in the keynote. Essentially, we're moving into this realm of Agentic AI. If you think about enterprise knowledge workers, we are on the cusp of giving every knowledge worker a team of superhuman capabilities embodied in AI agents that go and think and reason and act autonomously on their behalf. Imagine having a team of 20 people working for you around the clock day and night, doing work for you, helping you make decisions, helping you choose the best alternatives when you can't. And so the age of Agentic AI and this autonomous reasoning embodied in these software agents that will work on your behalf is huge. I think every enterprise is going to embrace some form of that. And then the other thing I'll lean into is Physical AI. The physical embodiment of AI technology in the physical world, in industrial settings, in robotics, autonomous vehicles. All of those endpoints become landing points for AI intelligence. Guess what? To train the next generation of autonomous vehicles, of robots, of humanoids, you need this kind of infrastructure that we've been talking about and which Equinix is helping us bring to market. Is the classical three computer problem. You build your model, you simulate it in Omniverse, you deploy it in Physical AI, solutions, robots, cars, humanoids, whatever it is. That's what I'm really excited about. We're so close and it's only going to speed up.

Savannah Peterson

>> I cannot wait to continue this conversation with you over time. I just got to ask, is that a GTC friendship bracelet that you have on right now?

>> Yeah. Actually, a friend of mine, she's a big Swifty. Is that okay to say?

Savannah Peterson

>> Yes.

>> She's got a bag of 50 of them. If you need the inside track on a Swifty GTC bracelet, just let me know, I'll get you one.

Savannah Peterson

>> Tony, little did I know. I feel like I need to step up my fashion game. I've got my Equinix pin. I definitely need a Swifty GTC bracelet.

>> Yeah. Absolutely.

Savannah Peterson

>> It's got to match my nails, right?

>> Yeah, exactly.

Savannah Peterson

>> Oh my goodness. Tony, thank you so much for taking the time.

>> Oh, thank you.

Savannah Peterson

>> Congratulations. Really exciting announcements with Equinix and frankly, one of the most exciting weeks of the year for you, I'm sure.

>> We're really excited and more to come, for sure.

Savannah Peterson

>> Awesome. Well, we can't wait to hear all about it next time you're on the show.

>> Thank you.

Savannah Peterson

>> Thank you all for tuning in to our exclusive coverage here from the Equinix Booth at GTC NVIDIA here in San Jose, California. My name's Savannah Peterson. You're watching theCUBE, the leading source for enterprise tech news.