What's the idea that led to 3DGS?
Our current chief technology officer, Jeb Flemming, founded 3DGS initially as a life sciences oriented company. The original name was Life BioScience Inc. Jeb realized that certain photosensitive glass-ceramics can be patterned and etched into fine 3D geometric structures and, subsequently, started the company in 2009. After several years of pathfinding, he realized the true potential of the technology may be better suited for manufacturing high frequency RF passive components. The company pivoted to this core business and changed the name in 2014 to 3D Glass Solutions Inc. (3DGS). I joined the company as CEO in 2017.
How was your glass integration technology developed?
Glass-ceramics were invented pre-World War II by researchers at Corning Inc. 3DGS’ current integration did not become feasible until Jeb Flemming came up with the idea to leverage the physical nature of the glass-ceramic structure to create low cost, highly precise 3D features using semiconductor fabrication techniques.
Unlike conventional glasses used in some semiconductor applications, 3DGS technology relies on manipulating the glass and ceramic states of the substrate. The manufacturing process doesn’t need expensive laser and etch systems to create 3D features. Moreover, 3DGS technology can create features of any shape and size—versus only circular holes from laser processing.
The ability to create these precisely controlled, different features, with full metal patterning into circuit elements, allows unparalleled levels of integration at great economies of scale.
How is this integration approach different and better than other heterogeneous packaging technologies?
This approach is unique first in the use of very low loss glass substrate material as the base. Wherever the design requires a 3D feature, we use lithography techniques to first convert the glass region to a ceramic, then etch the ceramic away in dilute acid. The result is micron-level accuracy, chemical machining that then can be further metallized with circuitry.
We use this technique to create precise passive elements like inductors, capacitors, resistors, antennas and filters that are fully integrated in the glass substrate. The devices can be stand-alone, surface mounted into SiP modules or active ICs may be mounted directly on the substrate by flip chip or wire bonding.
Compared to LTCC or laminate approaches, 3DGS is economically competitive while delivering higher performance at microwave frequencies, resulting in very low loss system performance and high component integration.
Highlight the performance capabilities of the 3DGS process (e.g., frequency coverage, power handling and thermal).
3DGS technology operates from DC to approximately 300 GHz and is best suited for applications in RF above 1 GHz. The microwave loss characteristics of the glass are very low, while the self-resonant frequency of the glass is much higher than conventionally used materials, resulting in high Q in the frequencies of 5G and beyond.
Additionally, since 3DGS technology can be used to create buried (under chip) antenna waveguides and filtering, the technology enables true antenna-in-package solutions that have ultra-low loss for applications like automotive radar and microwave backhaul transmission at E- and D-Band.
From a power handling perspective, there are two concerns: one being the ability to have thick Cu lines and features, and the second is thermal dissipation. While 3DGS can create very thick Cu traces and 3D routing, the management of heat is done through precise locations of Cu vias for thermal drain. Our technology provides the designer with the flexibility to place any number of Cu filled vias in different substrate locations, even of multiple diameters, at the same time.
As 3D glass has wide applicability, from components to highly integrated subsystems, what markets are you targeting and with what types of products?
As a small but growing company, the key here is focus. While 3DGS’ technology is a “platform” upon which many different solutions can be built, our focus over the next several years is 5G RF discrete and IPDs, mmWave filters, mmWave antennas and antenna-in-package and full integration of all components into heterogeneous solutions. While this still seems like a lot, the common thread is that the manufacturing process and machines are the same for all the aforementioned products.
We have strategic investors who are also customers and collaborators, so all of our products are designed and built with a specific customer application input/output.
Describe the complexity of the manufacturing process. Do you have the production capacity to support your near-term market opportunities?
3DGS’ manufacturing process can best be described as a combination of 20-year-old semiconductor fab technology combined with modern WLCSP back-end technology. The front-end of our process is relatively simple and done in batch. The back-end involves thin film metallization, electro/electroless plating, contact lithography and final test/pack.
We have around a ~35,000 square-foot (350 square-meter) factory in Albuquerque, N. M., which is a very low cost footprint. We leverage slightly used and new equipment in our processing, since the technology node is quite old by semiconductor standards.
Our capacity is growing and we can accommodate our growth internally over the next several years. As our volume needs grow, we can employ the traditional outsourced semiconductor assembly and test (OSAT) model where appropriate for the back-end of the process. Any typical WLCSP OSAT can be used. The front-end of technology is more proprietary, so outsourcing may be done on a licensing collaboration model.
As you plan for the next five years, where would you like to position 3DGS—i.e., catalog component supplier; partner with systems companies developing custom, highly integrated subsystems; open foundry?
We definitely will be offering a certain amount of catalog components, but the real core growth is likely to be in build-to-print products.
Our process cycle time can be very fast, and we have partnered with Cadence-AWR to develop an RF PDK. The PDK is deployed to our customer via seats from Cadence-AWR and we provide a free key to the software.
RF designers at the customer location can design their components with our background IP, and we work with them to finalize the design and build a hybrid foundry model. The differentiation in our technology is that our process flow is simple enough to allow fast cycle time of even very complex integrated components, coupled with a high degree of agreement between design simulation and actual product performance.
This is a game changer in IPD-type products and allows the customers to reduce their time to market without sacrificing space and loss of using conventional discrete components for high frequency applications. Even if the customer wants to start with the discrete component solution, we have that solution ready as well.
How are you funded and where are you in your business plan?
We are a growth stage, venture-backed company. Our investors, however, are mostly strategic, including Nagase Group, Lockheed Martin, Murata Manufacturing and Corning. The next 12 to 24 months is about ramping our design wins and expanding our customer base while we add manufacturing capacity and hire additional engineers and technicians.
What have you learned about getting a new technology commercialized and adopted?
When I joined 3DGS in 2017, it was after an already long career in both RF and the semiconductor space. I’ve seen a lot of successes—and my share of failures, too. Like many companies with a new and potentially disruptive technology, 3DGS’ initial focus was on building complex and custom solutions versus competing in discrete, commodity solutions. I did not think this was the way to grow. We were also doing too many things for too many different customers.
First, we had to stop and realize that real disruption usually occurs from the bottom and before we can move to the top, we have to show the customer engineers that our inherent discrete solutions are significantly better than conventional solutions at the equivalent economic price point. Moreover, there needs to be a lot of data generated on a one-to-one basis to convince customers that the new technology both adds value and provides comparable or better reliability.
We also had to have a strong software enabled PDK toolkit so that our customers could design and simulate on their own in their own application environment.
Finally, we have to deliver prototypes and samples quickly so that we establish feed-forward/feedback relationships and create successful customer collaborations.
Doing less sometimes equals doing more in the long run. Find your partners early and, hopefully, bring them in as investors so they have skin in the game. Don’t build new technology in a silo.
Tell us about your background and what led you to this CEO role.
I have a pretty diverse background. First, I’m a materials engineer with a specialty in ceramics and glass. Second, I started at the bottom and gained a lot of experience in manufacturing, R&D and design. Third, I have a lot of successful experience and scars in business development, strategy and team leadership.
I was fortunate to have an early start at Motorola where I gained significant expertise in Six Sigma and lean. I joined a team that started ChipPAC as a spin-off from Hyundai Semiconductor and led the efforts to establish business with the burgening fabless companies at the early days of BGA and chip scale packaging. I lived in Asia for 10 years serving as corporate vice president, leading Amkor’s business growth in both Japan and the rest of Asia, helping form J-Devices and leading buy-outs of several Toshiba factories. Most recently, I led the electronic business unit and corporate business development teams at Henkel Corporation. I also serve as a member of the board of directors for Intevac Corporation.
3DGS has fantastic technology, a strong and growing team of entrepreneurial-minded employees and a low cost, scalable process. The technology has a real use and can be very disruptive. These were the key factors that attracted me to become the CEO of 3DGS, and I’m lucky to be leading such a great team.