What is the focus of your technology at Eliyan?

Eliyan was established to facilitate and accelerate chip systems. Our vision is to create the ultimate chiplet system. A chiplet represents a miniature integrated circuit (IC) with a distinct set of functions. Its purpose is to merge seamlessly with other chiplets on an interposer within a single package. Especially in this era where we're moving from SoC – which is system on chip – to system on package. Because of the limitations of Moore's Law, where we are reaching the limits of what can be achieved on chips, chiplet interconnectivity is the dominant enabler of creating the bandwidth and memory performance we will need for the future. And we're at the center of it.

What are the core problems that Eliyan’s technology solves for customers?

In the chip industry, customers seek more and more performance out of semiconductor chips. The past half-decade or so has seen a significant shift in the semiconductor industry. For many years Moore's Law – which forecasts a doubling in chip performance every two years – was the steady benchmark of performance growth, but this has begun to slow down considerably. Instead of doubling transistor count and performance every two years, we started seeing incremental gains of 50%, 30%, and even 20% with each new process technology. This slowdown put a real cap on the potential performance of individual chips. On top of that, the bandwidth between these chips has been very limited.

The performance of computer chips has gone up by five orders of magnitude, and the bandwidth between them has only gone up, maybe 100X. That created yet another issue what people call a memory wall or I/O wall. If you want to exit the package from one chip and go to the other one, there are a lot of limitations. You cannot build packages to be tinier and tinier and so forth.

With chiplets, we can address these problems. By combining multiple processors as chiplets within a single package, we can double compute power without pushing miniaturization limits. This modular approach also allows us to integrate dedicated memory chiplets, shattering the bandwidth bottleneck that once hampered communication. Chiplets help address the chip industry's biggest roadblocks in recent years and that’s why we believe it is the future of performance in this industry.

Is there something new that differentiates Eliyan from competitors?

The chiplet revolution has been simmering for several years, but it’s gaining momentum now. We were among the first companies to recognize its potential to overcome bandwidth constraints and dive headfirst into developing chiplet-based solutions. Our proprietary chiplet interconnect technology, pioneered over seven years ago, is a cornerstone of those solutions. Eliyan’s NuLink™ PHY technology can be optimized for high-speed serial die-to-die connections using any industry standard or custom interconnect scheme. It can deliver outstanding performance at low power in advanced packaging or standard packaging using a wide range of bump pitches. This head start and our deep experience gives us a clear edge in the field, solidifying our position as a leading innovator in chiplet architecture.

What is the long-term vision for Eliyan?

Our ultimate goal is to facilitate the full adoption of chiplets and democratize the use of chips. Our strategy to reach this goal is focused on building a foundational infrastructure of differentiating chiplet technologies and then establishing and expanding a portfolio of products and solutions around that.

If you look at a successful company like Broadcom, they started with high in-demand connectivity technologies that were superior in the industry. Having that differentiation, they were able to build other chips, compute chips, switch chips, and so forth that benefited from that superior bandwidth and lower power of their connectivity technologies. They continue to lead that market. At Eliyan we put our focus on not the conventional systems with conventional chips, but on chiplet systems in a package with chiplets. Because the future will be dominated by the systems of chiplets in a package. We deliver the highest performance, lowest power, chiplet connectivity that is 2-3 years ahead of the industry.

As with Broadcom, with Eliyan’s technology, once we have established our differentiations, we can expand into different products that will be inherently differentiated. We plan to be the Broadcom of tomorrow, delivering the ultimate chiplet solutions for our customers.

Are there any exciting things happening in the business right now?

We are seeing a lot of growth in customer interest, which is very exciting. Companies who want to adopt chiplets have very limited solutions today. Ideas that we pitched three years ago that were received with hesitation, now have some major companies circling back around. A lot of companies in the market are coming to us so we’re realizing we need to expand as fast as we can to address this demand.

We’re also focused on solidifying a strong competitive advantage in our technology, which is also exciting to us. We have thought about all the products that people will eventually need to build and create a powerful strong patent portfolio around those solutions. Having a strong patents portfolio makes the implementation of our products exclusive to us and puts us in a nice position in the market for the long run.

The chiplet revolution has been simmering for several years, but it’s gaining momentum now. We were among the first companies to recognize its potential to overcome bandwidth constraints and dive headfirst into developing chiplet-based solutions.

‍

Could you give us just a brief version of your professional background, and what brought you to founding the company?

My background is in high-speed connectivity. In the early '90s, I started in the RF domain, then wireline communication caught my attention. I did my PhD in that domain in wireline communication at Stanford. I invented the first type of SerDes, called PAM4, to overcome the bandwidth limitations of silicon and channels.

By 2006, I started a company called Aquantia, focused on Ethernet connectivity. We led that market together with Broadcom. In later years, we also developed many new Ethernet technologies, including one that could run much faster, 5 gig and beyond over the existing enterprise CAT5e cable infrastructure. That technology became an IEEE standard in 2015. Following that work, we also proposed the solution for Ethernet in the car for 10 gig and beyond, 10 gig and 100 gig.

My focus started with SerDes technology, then Ethernet connectivity, and now, the chiplet interconnect at Eliyan. Different types of connectivity technologies from hundreds of meters to a few millimeters. But my personal goal has always been how to create the highest-performance connectivity technology.

From your experience, how could an investor best partner with a founder and a startup company to support them and help them grow?

Investors of course are key for connections and influence. Their networks are typically very strong and can be helpful for a newer company trying to establish itself.

When you start discussing a disruptive technology or solution with a potential customer, especially with big companies, it can be challenging. But if you start these conversations through your investors, there is more possibility to build connections at a higher level with people who see the value of your technology and how it will benefit their company. Getting to the right people the first time helps us to get a real seat at the table and accelerates the process considerably.

It is also important to have investors who understand your business and are committed to your success for the long term. Fortunately, we have investors and board members who believe in us and don’t quit when there are setbacks. To have that trust and support is paramount.

What advice do you have for other founders starting a new company?

Success starts with a burning belief in your vision. Even when the road gets tough, unwavering faith in your vision will keep you going. There will be moments of doubt but remember why you started and how your solution addresses a real need. Always make sure you are solving a big enough need with a big enough differentiation.

When building your vision, create something truly visionary. This requires genuine hard work – deep market research and actively listening to customers to understand their real challenges. Don't just fix problems you think exist; anticipate future needs and develop solutions for them before they become mainstream.

At the same time, you can’t drink your own Kool-Aid; constantly validate your idea against the needs of the market. Building something people don't need is a recipe for failure.