Skip to playerSkip to main content
  • 11 hours ago
Transcript
00:00Welcome to our Bloomberg television and radio audiences worldwide.
00:03In Washington, D.C., one of the central players in global AI infrastructure build-out
00:08is speaking to and about its most important and biggest customer,
00:13that is Dell and the United States federal government.
00:16Delighted to say that Dell Technologies Chairman and CEO Michael Dell
00:20joins us alongside the Department of Energy Undersecretary for Science, Dr. Dario Gill.
00:25And Michael, a lot of the themes that you're trying to get across,
00:28and I don't know what conclusions you've reached in the last hour with your colleagues sat next to you,
00:33is that either government is getting it right in the United States or they're not,
00:38in where they're allocating resource, where they're deploying policy
00:41to build critical national AI infrastructure.
00:45What are the conclusions you've reached today?
00:49We're seeing in government a number of the projects advance from the pilot phase to production phase.
00:58And certainly, you know, Dario can speak to the ambitious initiatives,
01:04but, you know, we are, you know, we have built a platform for open, scalable systems
01:14to enable all the great sciences occurring in the national labs and beyond.
01:19Certainly, we have a fabulous example in the Doudna cluster that we're beginning to deploy.
01:25And all of this is about accelerating science, you know, scientific discovery and progress.
01:34And certainly, you know, that's a huge priority.
01:37So great to be here with our largest customer.
01:41And, you know, we're very mission focused and making sure our solutions are enabling the mission wherever it occurs,
01:51whether that's in space, you know, out in the field and everywhere else.
01:59So, there are case studies, right?
02:01We could talk, for example, about the progress you've made with NERSC.
02:05But zooming out for a moment, if you will, from the DOE's perspective,
02:11why is AI infrastructure now such a core part of national security and energy security implications?
02:19Well, because we're in the midst of a computing revolution and AI is at the heart of that revolution
02:26and it is going to transform how science and engineering are practiced in our nation.
02:31The Monday before Thanksgiving, President Trump signed an executive order to launch the Genesis mission.
02:39And within that mission, we seek to harness this computing revolution to double the productivity and impact of America's trillion
02:48-dollar-a-year R&D ecosystem.
02:51And science and engineering are at the heart of our progress and capabilities in national security, in discovery science, in
02:57energy, in health.
02:59And the example of the partnership that we have with the doubt in the system in NERSC, in one of
03:05our 17 national laboratories,
03:07exemplifies the power of using AI to accelerate science.
03:12Dr. Gil, is Dell and Mr. Dell sat next to you delivering supercomputers to the timeline you'd expect?
03:20Yes.
03:21So, for example, in that system that we're standing up,
03:24we already have an early access program where we stood up the first generation of nodes of the NVIDIA GPUs.
03:31The system is going to be operational for the science community mid-2027.
03:37So, I think one of the characteristics we're seeing right now is that we used to have supercomputing projects that
03:43would take a decade to build.
03:44Now we're building these things in two years.
03:46And that's a big, big change in terms of the velocity with which it's moving.
03:52One of the things that the government is doing is, sorry, I had to interrupt there.
03:57No, please, please carry on.
04:00Leveraging the integrated, you know, rack-scalable systems that we're building for commercial and enterprise.
04:07And obviously, there are enormous build-outs in AI data centers, as you're well aware, Ed.
04:13And so, you know, kind of, again, leveraging that open, scalable architecture that we built
04:22will accelerate the government's progress in all of its AI infrastructure requirements.
04:28We're live on Bloomberg Television and on Bloomberg Radio.
04:32Around the world, we're speaking to Michael Dell and Department of Energy Undersecretary for Science, Dr. Dario Gill.
04:38Michael, the Pentagon recently designated Anthropic as a potential supply chain risk.
04:44Your assessment of that decision and the scenario that's playing out there, please.
04:51You know, I try not to assess that decision.
04:56You know, I think, you know, we're kind of a foundational provider of the infrastructure.
05:10What I can say from our standpoint is, obviously, we have various controls and systems to make sure we're selling
05:19to the right, you know, to the authorized users.
05:25But we, you know, I don't think a company can dictate to a sovereign government what it does with its
05:32tools.
05:33I just don't think that's a workable model.
05:36Dr. Gill, I ask that question because right now, you know, your department, every department, has to balance what our
05:42national security concerns and interests with very real cyber threats to U.S. infrastructure.
05:49How do you do that?
05:51You know, keep working with private enterprise American companies, but also achieve the goals that this administration wants to in
05:58this period?
06:01So, look, I mean, specifically, if you look at cyber threats and the overall security posture on these technologies and
06:10being very aware that our adversaries do not, you know, wish as well,
06:14the way you approach it is to, from day one, build a culture and a technical architecture where cybersecurity and
06:23security overall are an integral part of everything that we're building, right, from day one.
06:28So, this is not something that is an afterthought where you're building an architecture or implementing these systems.
06:35And frankly, one of the things, for example, that we're seeing that is an evolution is that AI-driven cyber
06:41attacks are increasing in its power.
06:44And the only way we're going to be able to defend it is with AI-driven cyber defenses.
06:48So, for example, categories like red teaming and blue teaming, what I am seeing now from our teams is that
06:55you're seeing those automated and you're seeing a red team and a blue team that is driven by AI agents
07:02attacking and defending each other.
07:04And then that informs the overall sort of like cyber architecture that we've implemented.
07:08So, this is one of the defining challenges in the deployment of technology, the overall cyber posture.
07:16Michael, there's great demand to understand what's happening in global supply chains right now, particularly with, you know, it's not
07:23necessarily delays in specific data centers, but shifting priorities.
07:27Where a data center needs to go first geographically.
07:30What are you seeing right now on particular bottlenecks or where, you know, from a CapEx deployment standpoint, the priority
07:36really is?
07:39Well, certainly, you know, as we've talked about in our latest earnings, the demand is ahead of the supply.
07:49And the constraints are the ones that we all know about.
07:54It's memory silicon, it's advanced node semiconductors, and power and the infrastructure to build out the power.
08:03Those are definitely all constrained to one degree or another.
08:10The good news is we're responding to the challenge and we have an incredible supply chain.
08:16And, you know, we're able to provide an enormous amount of this to our customers because we've got our arms
08:26around a lot of supply and a supply chain that is globally distributed, resilient and works well, maybe even better
08:35when there are challenges.
08:37Under Secretary, you know, with the war in Iran, there is a focus on energy prices, on oil.
08:42Is there anything to extrapolate out from that longer term on your departments and this country's ability to secure energy
08:52supply for the data center capacity that is being built?
08:57Energy is an absolute top priority for the administration, for President Trump, for Secretary Wright.
09:04And energy is the other side of the coin of the AI computing revolution.
09:10So we need a world of more energy and, you know, more abundant energy and secure energy systems.
09:17But look, there is no way for us to deliver on the ambitions that we have as a country for
09:24AI and for additional technologies and the reindustrialization of America without a growing energy and affordable energy supply.
09:34So this is an absolute top priority.
09:36And I'll give as an example.
09:38I mean, we're talking about oil and gas, but also in the horizon is a renaissance of nuclear power that
09:44we're very driven.
09:45If you look at even beyond is our commitment, our investments towards realizing fusion energy.
09:50So we need to invest across all the different horizons to realize the full potential of America.
09:55America.
Comments

Recommended