The semiconductor industry dynamics are changing. General-purpose, centralized computers fueled its growth for decades with advanced processing power. Now, with that growth slowing, the compute power is moving to the periphery, creating applications in healthcare and making open-source tools more viable. Geographic locations may no longer be a consideration so chip designs can now be done everywhere.
Lucio Lanza, Managing Director of Lanza techVentures, led three seasoned executives on a tour of the current semiconductor industry landscape during the recent GSA Silicon Leadership Summit. Their broad conversation focused on the democratization of chip design and the impact it could have on healthcare.
Joining Lucio were:
- Max Mirgoli is Executive Vice President of Worldwide Partnerships at IMEC and a member of the IMEC Executive Board.
- Mike Wishart is Chief Executive Officer, Co-Founder and Director of Efabless, a creation platform for ICs and related electronics through open-source and community models to make design and prototyping streamlined, simple and inexpensive.
- Francis Ho is Partner at Walden Catalyst Ventures, and his expertise in science and mathematics and his deep operating experience led him to a career in venture capital
What follows is a condensed version of their conversation.
Lucio: With Moore’s Law slowing down and the strong second derivative becoming negative, the structure of the industry has changed. Everyone assumed the semiconductor industry would continue at its pace of integration. In reality, that is not happening at all. That’s not the trend.
Since centralized intelligence cannot move as fast, intelligence is at the periphery and computing intelligence in the periphery is exploding. As that happens, the community of designers explodes. When the community of designers explodes, the right kind of tools need to be available through an easy-to-access tool system and that is going to be a different structure. It is called democratizing design.
It will not be specialized or something only smart engineers can do. Once that happens, the investment community’s support of this new structure will change and the expansion will be broader.
What in theory was going to be a reduction of investment is an explosion of new investments.
Max, can you talk about the expansion of the investment community.
Max: By the virtue of my position in our organization, I get asked the questions, “What’s going on in the industry? Why is semiconductor so hot?” Economists will say there was a pent-up demand with COVID and consumers have disposable income. That’s a good explanation.
There’s a new breed of investors. There was a time when investors in deep tech were people like you, Lucio, and now people like Francis from Stanford University with a PhD investing in startups solving hard to imagine challenges. There was a time that quantum computing was only for the big companies. Now there are unicorns that are AI companies and Bitcoin mining. All of this is semiconductors.
Lastly, one thing that has happened, perhaps the big iceberg, is the democratization of chip design and the lowering of the barrier to entry. Apple ads were about the product. They now are about the chip. Designing a bespoke customized chip can differentiate the company.
And, then there is Efabless that allows the barrier to entry to get even lower to the $10,000 level. Google Cloud offers an ADA tool so entrepreneurs with even small money can access powerful design capabilities and don’t need to go to traditional venture capitalists. It’s amazing how much talent can now come to the table, and how this talent can generate new IP, new designs and it’s mushrooming. Add to it the fact that every government is looking at the geopolitics, supply chain resiliency and throwing money on the table.
Everyone wants to control their destiny in the supply chain, but also the innovation. The digital dollar is no longer a question, it’s a reality. It’s not if, it’s when and these changes are all coming from semiconductors.
The complexity and how many chips are needed in every traffic light and signal. Look at Tesla and why it is unique. It is because of all the hardware and software Tesla put together. It’s a combination of renaissance in many areas of our living.
Look at the democratization of healthcare and it’s coming from chips. This democratization of chips and the hardware/software combination is opening and creating a better standard of the living and solving many problems of the society, such as carbon capture and climate. Much of it can be controlled through technology.
Lucio: Let’s hear Mike’s view on the democratization of design, and how meaningful it is, and impactful it is going to be.
Mike: The ability to democratize design is opening innovation. The metaverse combined with open source and decentralizing platforms like ours enable people who perhaps have never met before come together to collaborate around projects. People with needs can meet people with skills to make products and chips that are more precisely tailored to the specific needs of an application. Designs can be shared and environments emerge that simplify design. The result is more designs and more innovation.
A typical question: “What chips can we make to enable my product, solve a unique problem or even to solve an existing problem in a more elegant way?” With this question, the concept of special-purpose silicon is born. Special purpose silicon has been around forever, but it previously existed for only a relatively few applications due to the cost and complexity of design.
With the advent of markets like IoT and machine learning at the edge, special-purpose silicon is increasingly needed to serve lots and lots of emerging markets. It means that the number of designs and designers has to explode. The intimacy of the application to the semiconductor design has to go up. The traditional way where someone who doesn’t know the silicon side is trying to communicate with a limited number of organization or people who do won’t work at scale. There are simply too many projects and the returns are too speculative.
That means a move from a world of 100,000 or so silicon designers to every software and hardware engineer being able to design. Efabless is a platform to handle everything from design to the actual packaged silicon, and make it easy, fast and simple. Around this platform and open source a community is built, a model that other industries recognize. The community becomes supportive of one another and the output of this is shareable designs.
Efabless works with Google. We taped out around 250 designs through community designs in the last nine months, perhaps 225 of which are shareable open-source designs, many designed with open-source EDA tools, including commercial EDA environments.
Designers should be able to design where they’re most comfortable.
Lucio: It is not a revolution. It’s not changing. Everything is expanding.
Mike Wishart: It’s a market expansion.
Lucio: Exactly. Francis, let’s move in a different direction and look at application that might be meaningful for society. For example, how are we going to see the healthcare application change? After COVID, people love having remote connection and remotely taking care of their health. Is this something that can happen to only the existing and well-developed communities in the United States or Europe, or is something’s going to happen for mankind everywhere?
Francis: I’m originally a physicist and worked on semiconductors such as analog chips and optics, but began going into healthcare about 10 years ago and that’s one of my focus areas today, together with web3 and my original sector of semiconductors and deep tech. I also resonate with what Mike said, and it’s not only about scalability of the core technology ––the chips, more bits, more throughput, lower cost. The other scalability is people, because nothing works without people.
What Efabless is doing is scalability of developers and users. At Walden Catalyst, we think about investing in core technologies that have to do with scalability of the technology and the key people, and then figuring out new use cases, new directions. One is healthcare.
We understand more than ever how important healthcare is. It’s obvious, but it’s hitting us in the face. You can’t survive, you can’t be here without the right things happening. A part of it is that we now know so much about the underlying biology, more than we did even 10 years ago and we have a shot at getting ahead of it.
An example is COVID. Within months after finding it, there were candidates being tested. That was not an accident. It was the result of decades of hard work and scientific investment. So we are accelerating the discovery of new therapeutics to help the whole world, and semiconductors have a key role in that.
Semiconductors are often involved in two ways. First, semiconductors on the supply chain side, such as equipment or EDA, is about moving atoms around precisely and organizing it using systems and designing it. That is useful to healthcare because new kinds of sensors use extremely precise technologies and often borrow from semiconductors. It’s not a surprise that many of the biggest companies in the semiconductor supply chain are actively looking at opportunities related to healthcare and life sciences because they see an overlap with their core skills, in the physics, chemistry, material science and software organizing the complexity.
The second way that semiconductors are involved relate to the electronics devices. Computers are gathering information, transmitting and processing. It’s allowing us to understand our world in digital form and do things about it and have insights.
Those two trends are directly related to new therapeutics and healthcare today. With COVID, those two trends made it possible for us to figure out the right vaccine, because we knew the shape of the virus, the shape of the spike protein.
Lucio: Once you peripheralize patient care, how do you protect that information? At this point, you have something personal, and something that, once lost, cannot be recalled. How do you work to make that safe?
Francis: That’s something we have been looking at through different investments and it’s a complicated question. There are many different legal issues and jurisdictions in Europe, the U.S., etc. Beneath that are industry conduct issues. In the United States, several stakeholder groups including payers, providers, patients, pharma, and medical devices have access to data, and each stakeholder has different incentives. The patient has the right ultimately to control that data. Complicated rules, regulations and practices control that.
New technologies called confidential computing and homomorphic encryption are making it possible to collaborate without leaking information. A new technology called Zero Knowledge Proof is another way to protect the data.
Companies are trying to take the data, use compute infrastructure that’s safe, and innovate on top of that to analyze patterns to generate insights. All this is happening in parallel, but we need to be patient because change comes slowly in this industry.
Lucio: Do you think they are going to help making the world fairer and more uniform?
Mike: By democratizing things and allowing the long tail to happen, solutions can be created to solve individual problems with individual answers. Platforms are needed where the economics change and the interaction between the people with the idea and the people who can solve it ensures everyone can contribute. Let’s call it micro market solutions that collectively are enormous. They would not see the light of day in a traditional system.
Francis: I want to talk about decentralization that is about giving people the power and not be subject to censorship or centralized control. That’s part of what’s driving web3 and a desire to make it democratized with the freedom to experiment and try things. That part of blockchain is interesting and there are many experiments that try to apply that to healthcare.
Max: The U.S. is a capitalistic society. Investors like Francis who are investing and making money in science that eventually will make lives better, richer and healthier is fantastic. I’m happy that they are. It’s not a mutually exclusive concept. If you look at the life expectancy of a woman in India is somewhere in the 30s. That’s a travesty. Look at what silicon did for computing.
Today, we have an HPC in our pockets and none of us understands this mobile device. Look at that technology and what you can harness from that technology to bring diagnostics to a village in India. Or bring a hospital in the form of a chip. It’s parallelization that the semiconductor industry started.
Lucio Lanza: I can see that we’re going to have support for all human beings. That is happening and I’m looking forward to it.
For more information:
Nanette Collins for Lanza techVentures