This conversation examines chiplet-driven data center architectures and ai factories. Mike Han of Silicon Box, chief revenue officer, joins theCUBE Research hosts John Furrier and Dave Vellante to discuss the company's origin, technical approach and manufacturing strategy. Han outlines the focus on chiplet architectures and panel-scale packaging and explains how these capabilities enable high-density heterogeneous integration; they describe manufacturing in Singapore and planned expansion in Italy and explain how these innovations support artificial intelligence applications in data centers, automotive, robotics and broader semiconductor markets.
Key takeaways include that chiplet-based designs and panel-scale packaging reduce interconnect costs and enable substantially higher density, offering a 4–6x productive-area advantage versus traditional wafer-scale approaches. Han reports strong cross-market demand beyond AI, including mobile and electric vehicle, EV platforms. theCUBE analysts John Furrier and Dave Vellante underscore supply constraints and the talent mix required such as electrical engineers, physicists and materials scientists, and they highlight the strategic impact of the Italy factory expansion for manufacturing capacity and supply chain resilience.
Forgot Password
Almost there!
We just sent you a verification email. Please verify your account to gain access to
theCUBE + NYSE Wired: AI Factories - Data Centers of the Future. If you don’t think you received an email check your
spam folder.
Sign in to AI Factories - Data Centers of the Future.
In order to sign in, enter the email address you used to registered for the event. Once completed, you will receive an email with a verification link. Open the link to automatically sign into the site.
Register for AI Factories - Data Centers of the Future
Please fill out the information below. You will receive an email with a verification link confirming your registration. Click the link to automatically sign into the site.
You’re almost there!
We just sent you a verification email. Please click the verification button in the email. Once your email address is verified, you will have full access to all event content for AI Factories - Data Centers of the Future.
I want my badge and interests to be visible to all attendees.
Checking this box will display your presense on the attendees list, view your profile and allow other attendees to contact you via 1-1 chat. Read the Privacy Policy. At any time, you can choose to disable this preference.
Select your Interests!
add
Upload your photo
Uploading..
OR
Connect via Twitter
Connect via Linkedin
EDIT PASSWORD
Share
Forgot Password
Almost there!
We just sent you a verification email. Please verify your account to gain access to
theCUBE + NYSE Wired: AI Factories - Data Centers of the Future. If you don’t think you received an email check your
spam folder.
Sign in to AI Factories - Data Centers of the Future.
In order to sign in, enter the email address you used to registered for the event. Once completed, you will receive an email with a verification link. Open the link to automatically sign into the site.
Sign in to gain access to theCUBE + NYSE Wired: AI Factories - Data Centers of the Future
Please sign in with LinkedIn to continue to theCUBE + NYSE Wired: AI Factories - Data Centers of the Future. Signing in with LinkedIn ensures a professional environment.
Are you sure you want to remove access rights for this user?
Details
Manage Access
email address
Community Invitation
Mike Han, SILICON BOX
This conversation examines chiplet-driven data center architectures and ai
factories. Mike Han of Silicon Box, chief revenue officer, joins theCUBE
Research hosts John Furrier and Dave Vellante to discuss the company's origin,
technical approach and manufacturing strategy. Han outlines the focus on chiplet
architectures and panel-scale packaging and explains how these capabilities
enable high-density heterogeneous integration; they describe manufacturing in
Singapore and planned expansion in Italy and explain how these innovations
support artificial intelligence applications in data centers, automotive,
robotics and broader semiconductor markets. Key takeaways include that
chiplet-based designs and panel-scale packaging reduce interconnect costs and
enable substantially higher density, offering a 4–6x productive-area advantage
versus traditional wafer-scale approaches. Han reports strong cross-market
demand beyond AI, including mobile and electric vehicle, EV platforms. theCUBE
analysts John Furrier and Dave Vellante underscore supply constraints and the
talent mix required such as electrical engineers, physicists and materials
scientists, and they highlight the strategic impact of the Italy factory
expansion for manufacturing capacity and supply chain resilience.
In this interview from theCUBE + NYSE Wired: AI Factories - Data Centers of the Future, Mike Han, chief revenue officer of Silicon Box, joins theCUBE's John Furrier to discuss how chiplet architecture is powering the next wave of AI infrastructure. Han traces Silicon Box's founding story — built in 2021 by three semiconductor veterans who recognized the limits of Moore's Law and bet on chiplets as the architecture the industry would eventually require. He explains how the company's large panel-scale manufacturing, operating at four to six times the output of ...Read more
exploreKeep Exploring
What is Silicon Box's core technology and what makes it unique (its "secret sauce")?add
What does "panel size" mean in semiconductor packaging, and how does a larger panel size create value or advantage—particularly for manufacturing larger AI and advanced automotive chips and supporting scale-intensive applications like robotics?add
What are the company's next steps and overall roadmap or vision for the Silicon Box?add
>> ...Palo Alto studio connecting Silicon Valley and Wall Street. I'm John Furrier to be here with Dave Vellante, my co-host. Hello, I'm John Furrier, host of theCUBE here in theCUBE's NYSE studio. Of course, we have our Palo Alto studio connecting Silicon Valley to Wall Street. This is our AI Factory series, part of our NYSE Wired program and community. Again, global systems are developing. You're seeing the AI factories deployed in large centers at the edge soon. More and more distributed intelligent computing. Mike Han is here, CRO of Silicon Box, a very successful company building out the chips and the systems needed for all kinds of computing. All kinds of customers, anyone who's building systems as a customer, a big facility in Singapore pumping out lots of supply. Mike, thank you for coming on theCUBE. Appreciate it.
Mike Han
>> Thank you.
John Furrier
>> So take us through the Silicon Box founding story and vision.
Mike Han
>> So Silicon Box is founded based on, so there was these three veteran industry folks that really wanted to think about a world beyond Moore's Law. So then they had this idea of manufacturing with this concept of chiplets, which was different from how people had traditionally been manufacturing these chips. So really Silicon Box was founded on this concept of enabling this chiplet architecture. And then it was a great timing when they founded this company because all of these things like AI really require chiplet architectures to work. So it was a great thing where this technology that these three veterans had been championing for such a long time suddenly became the technology that was necessary for stuff like AI.
John Furrier
>> And obviously, the market demands. You guys have massive production. Congratulations.
Mike Han
>> Thank you.
John Furrier
>> A rapidly growing semiconductor company. What's the core technology? Explain the secret sauce without getting in trouble.
Mike Han
>> Well, we can't share the secrets.
John Furrier
>> No one's going to reverse engineer. Come on, share. No, talk about the core technology. You guys have the fabrication capability. Just take us through some of the high-level points.
Mike Han
>> Well, so what's unique about Silicon Box is that we can enable course interconnection. So integration of different chip components, add course pitch to also find pitch so at high density as well. So that's the technology that's necessary for AI chips, for example. And then we also manufacture that at a large panel scale. So two different things that make us unique. So what we're really doing is making what's traditionally very expensive in interconnecting different chiplets, especially at high density, we can make that a lot cheaper because we're doing it all at panel scale, this larger manufacturing format. That's why our factory is also really large scale technically because we're manufacturing at a 4, 6X scale as a traditional wafer scale foundry that would be doing this as well. So that's Silicon Box.
John Furrier
>> Chips in a box. Why is the timing right? By the way, the density is a huge factor in AI. It's a feature, not a bug. We're seeing that clearly in the systems. The packaging panels, again, another good call. Congratulations. Why is the timing so important now with AI?
Mike Han
>> Well, there's not that many companies that can do this. And then there's not much capacity available right now when there's a lot of demand for these devices right now. So I think we're in a good place because of that Silicon Box has been pretty successful even though we're a new company. So it's an exciting time.
John Furrier
>> When was the founding year? What year?
Mike Han
>> It was in 2021.
John Furrier
>> So recently. So it's new.
Mike Han
>> Yeah, relatively new.
John Furrier
>> Well, congratulations.
Mike Han
>> Thank you.
John Furrier
>> The edge has come up a lot. Density we're seeing happen. And when you start talking about density and panels, you go outside the thinking around normal AI conversations we're hearing. Oh yeah, gigawatt data centers, monster facilities, a lot of energy, big racks, power and cooling, liquid cooling. When you look at some of the inference and some of the things that are needed, it's not your standard GPU cluster. There's a lot of new super chips coming out that are microchips that need good form factor.
Mike Han
>> Yeah.
John Furrier
>> What's your perspective on this and how does Silicon Box help the evolution of where form factor, energy and networks exist?
Mike Han
>> So Silicon Box definitely helps reduce form factor as much as you can, even for systems that are technically much larger, which is what you really need for these types of technologies. And there's no real getting around it because these chips are now really big. So that's why they're difficult to manufacture and why Silicon Box is here, basically, to help.
John Furrier
>> The chiplet ecosystem, you mentioned difficulty. It's always been a hard thing. You guys are democratizing the chiplet ecosystem. How are you doing that? Give us some commentary on what that means. What does it mean to democratize the chiplet ecosystem?
Mike Han
>> Well, I mean, theoretically it's supposed to give customers a lot more flexibility to design chips without the traditional constraints that come with the massive investments that are traditionally required to develop new chips because you don't have to upgrade the entire system. You can theoretically mix and match different components of existing systems that work pretty good and then just focus the innovation on discrete parts of the system through this concept of chiplets.
John Furrier
>> So it's built for heterogeneous environments.
Mike Han
>> Yeah. I mean, so through heterogeneous integration we're able to jumpstart innovation because you can reduce the time and investment it takes to develop new chips.
John Furrier
>> Talk about the talent required to support these chiplets for you guys and for the market. What's the talent requirement? Is there certain skills?
Mike Han
>> Yeah, I mean-
John Furrier
>> What are some of those skills? Is it super semiconductor designers? What are some of the practitioners doing?
Mike Han
>> Yeah, electrical engineers, basic sciences, like physicists, material scientists, usually people in those departments are the ones that are involved with what we do.
John Furrier
>> Mike, it's great to have such great innovation. And we're here at the New York Stock Exchange. Of course, we have our Palo Altos studio. We see innovation on both coasts. Obviously, Singapore is now global. This building's a symbol of capitalism and money and capital markets. There's a lot of money in the market. CRO, what does that mean to you, does it keep you focused, execution? What's your plan? What are you optimizing for now? Obviously, there's demand. What's the focus?
Mike Han
>> Well, right now we're in an unusual situation where everyone wants what we're doing. And it's been lucky timing, I guess, because it's technology that these particular founders have been working on for a long time, but a few people can do it, and it's necessary for things everybody wants right now like advanced automotive and AI. So-
John Furrier
>> Robotics is big.
Mike Han
>> Yeah, robotics too.
John Furrier
>> That's a big market.
Mike Han
>> Yeah, anything that requires-
John Furrier
>> Physical AI....
Mike Han
>> a lot of scale is something that we can work on because the large panel size means that we support massive scale like robotics. So yeah, we're just trying to make sure that we can support all the demand.
John Furrier
>> Explain panel size, the panel packaging. I think that's a real key thing. What does that mean? For someone who's not in the market, what does that mean, panel size? What does that translate to for value?
Mike Han
>> So the panel size is the actual absolute productive area for what we're doing in terms of packaging. So the chips are actually integrated and then packaged on this larger panel format. That's a differentiator for what Silicon Box does compared to other packaging providers. And the fact that we can do it on a larger panel is great because a lot of these AI or advanced systems are usually a lot larger. So we're actually ideally positioned to support and manufacture larger chips like AI chips or advanced automotive chips.
John Furrier
>> And the custom silicon is a big trend. Integrating down into the silicon has been a huge advantage. You're seeing a lot of the smart engineers go there. I can imagine that the customer base, they're optimizing their software to run on the chips. Is that part of your workflow and engagement with customers?
Mike Han
>> What do you mean?
John Furrier
>> Well, do they come to you and say, "Here's our software. Build some panels for me."?
Mike Han
>> Yeah, definitely. I mean, they give us designs built on different types of software. I think that's what you mean.
John Furrier
>> I guess my final question for you is what's been the biggest change since founding? Obviously, the founder's got to be excited, their vision panning out. They made good bets. Product's good. Product market fit is there. Congratulations.
Mike Han
>> Thank you.
John Furrier
>> What's the feeling around the change? What's been the biggest change in the market? Has it been the tech? Has it been the demand, all the above? What have people missed, I guess? I guess maybe that's the question. In terms of the trends, why aren't anyone else doing this? I mean, you guys obviously hit this great opportunity. I mean, you said automotive, but there's plenty of other examples.
Mike Han
>> Yeah, yeah.
John Furrier
>> Digital technology with AI is dominating at the physical layer. What's been the big change in the marketplace?
Mike Han
>> I mean, what we've found that since we're offering such a fundamental technology is that a lot of other technologies beyond AI can benefit, and we've seen strong demand across a lot of different applications, not just AI. Right now a lot of people who are in chips keep talking about just AI or sometimes they'll talk about robots. But actually we're seeing pretty strong demand for what we're offering for even traditional applications like mobile, traditional automotive as well as EV. So I guess what we're seeing-
John Furrier
>> Hardware's back.
Mike Han
>> Yeah.
John Furrier
>> Hardware is hot, right?
Mike Han
>> So I think-
John Furrier
>> Software's easy. Hardware's hard.
Mike Han
>> What people are missing might be is that there's a general boom in innovation for chips across even traditional industries, not just AI, which is being obscured a bit by the whole AI boom.
John Furrier
>> Well, great to have you on.
Mike Han
>> Thank you.
John Furrier
>> What's next? Is there a vision for a roadmap? Can you share directionally what's next for a Silicon Box?
Mike Han
>> Well, we're building a second factory in Italy. So even though we're a relatively new company, we're already going to be doubling our capacity and building another factory in Italy. It's a 3.2 billion Euro investment, and we're getting a 1.3 billion Euro grant to support that investment, so it's an exciting development. We're going to be in production there 2028. So just trying to-
John Furrier
>> I mean, people want the chips. Software market's hot, open source is great, a lot of innovation.
Mike Han
>> Yeah, we're just trying to make sure that we capture all that demand right now.
John Furrier
>> All right. Call Mike. He's got the chips.
Mike Han
>> Yeah, definitely.
John Furrier
>> Supply chain is a differentiator, but also it's not just AI. Again, AI infrastructure, AI factory is only one dimension. The enablement that comes from these large-scale systems is going to spawn a lot of derivative value and value capture from application owners, system architects and hardware devices. So anything from small to big is all happening here in the AI wave and the infrastructure that's enabling it. Of course, we're doing our part here in theCUBE to bring you all the action. I'm John Furrier, the host. Thanks for watching.
John Furrier