Well, hyperscalers are pouring trillions into data center CapEx, but there's also a noticeable shift from wearable health monitors to industrial IOT. The semi industry is also facing an acceleration in demand to run AI computations directly on device without relying on cloud or. draining power grids. And at the same time, chipmakers are navigating renewed supply chain bottlenecks, not to mention stricter data privacy laws, as well as a re-rating of high-growth tech stocks across global exchanges. Well, here to break down the edge AI boom as what's happening across the semi-space and global tech capital flows is Humi Esaka, CEO of NYSC, listed Ambic Micro. Humi, great to have you here. Thank you so much for joining us. Well, Wall Street has been laser focused on hyperscale data centers as well as massive power grid constraints driven by cloud AI. So why is the broader tech industry now shifting toward on-device edge AI and can local processing truly relieve some of these data center bottlenecks that are out there?
Oh, definitely. Again, it's nice to see you and thank you for having me here. I think when we talked about a year ago, it was like very early inning of competition at the edge devices. Now that data center is growing, but more and more computation is required on the devices to offload some of the computation needs from data center. And also the functionality at each AI device is evolutionalizing month after month. So since the IPO in July 30th, artificial intelligence is entering our demand from our customers is like almost doubling. And if it's not much more than that, and we believe that computation on the edge will do two things. Number one, offload the data center computation. Second thing is that really, do more privacy and security on your devices so that if data center get some of the problem, your privacy and your data is going to be stored in your device so that there is less risk of your privacy and interpersonal data leak to somebody else.
And Fumi, while I have you here, demand for next-gen chips is accelerating faster than capacity can keep up. But where are the biggest manufacturing as well as supply chain constraints, and how is this threatening to slow down global AI hardware deployment right now?
Well, since last year, I believe that our demand grew almost twice as much, if it's not more, than last year. So demand constraint is definitely there. But for us, it's just because of the doubling demand and potentially going higher, that's where the demand constraint is. So we do still see the very healthy growth from quarter after quarter in the next couple of years. So we're not too concerned about demand constraint because of the fact that the chip is not as big as a data center chip. So we believe that again that things was that our demand skyrocketed and some of the demand constraint was because of the fact that our customer needed a product millions of pieces. within like a few months, which is a little bit physically difficult to do. That's where the constraint is. So we believe that as long as we work with customer very closely, we should be able to meet gross demand. And we believe that HAI market is really early inning in the game. And you will see so many demand is still coming out of a market and Amig will be able to support them.
And while I have you here, I do want to get your perspective on data, which is something that you mentioned at the top of the interview. And regulators as well as consumers are becoming cautious about sensitive personal data as well as health data. So this does apply to third-party cloud servers. So how critical is local on-device processing in overcoming regulatory hurdles as well as protecting user privacy when it comes to next-gen consumer hardware?
Just like I said, number one, we have a really good security system to make sure that unnecessary data will go up in the cloud. And our AI team has done a great development of AI software that we can compress the data and minimize data to be uploaded into the cloud. And that will enable us to minimize the risk of the data privacy. Definitely, I do not want anybody to know where I have been or what's my health parameters or those things. Definitely, those are the things that we can avoid by enabling our edge AI devices to do more scrutiny on privacy and security.
And Fumi, before I let you go, we are seeing a growing trend of companies, tech companies in particular, looking to do a list across U.S. and Asian exchanges to assess a regional liquidity pool. So what does this cross-border listing momentum actually tell you about how capital is being deployed and also valuing Western AI hardware innovations?
Definitely, we just did a dual listing at the Singapore. Definitely, a lot of engineering resources and also the market are global, especially HAI market. It's like coming out of everywhere. in the world. So we just wanted to make sure that we have enough capital to grow even outside of the US and also attract the talent and be close to our customers and supplier chain partners. So it was a great thing that we will be able to list both in New York and Singapore and we're very excited about it.
Hopefully we will have to leave it there for today, but it was wonderful talking to you. Thank you so much for joining us today and thank you so much for sharing your insights as well as your perspective.
Great, nice seeing you again.
