What’s really happening in the semiconductor industry? Q&A with Peter Suliga
November 10, 2021
Peter Suliga is an electrical engineer specializing in motion control solutions. With more than 15 years of experience in electronics, Peter is responsible for business development for the electronics industry in North America. He and his west-coast-based team work closely with HEIDENHAIN researchers in Europe to develop industry-leading solutions for suppliers across the semiconductor supply chain.
We sat down with Peter to get his take on the tumultuous, yet critical semiconductor sector.
Q: What impact does the electronics industry have?
A: The electronics industry is essentially everything that goes into the manufacturing of computer chips and printed circuit boards. It’s an industry that touches every facet of our lives. If we didn’t have a semiconductor industry, we would basically go back to the 1960s in terms of the everyday technologies we use. Every single LED light bulb you have at home is a result of the semiconductor industry and processes developed there.
We’re all connected today, everybody has a cell phone or a tablet. Just a few years ago, if you had home WiFi, it would have just a handful of devices connected. Now, you probably have 20 or 30 devices connected. We are living in the era of Internet of Things so you can find chips in just about everything.
Q: Where is the electronics industry heading?
A: It’s growing exponentially. So much so that Intel CEO, Patrick Gelsinger, recently made a bold prediction: In 2019 computer chips made up about four percent of a car’s bill of materials but by the end of 2030 it will look more like 20 percent. That’s just one example, but if I look at what they’re trying to do with computers in cars, I can clearly see why we need a lot more capacity and why the industry is going to be growing significantly into the foreseeable future.
Q: Can you explain the geographical breakdown in the electronics supply chain?
A: There are really two sides to it: one is the factories that make the chips and the other is made up of the companies who make chip production possible, capital equipment builders. If you think about the biggest suppliers of capital equipment, so-called tier one, we have Tokyo Electron in Japan, ASML in the U.S. & the Netherlands, and based in the U.S. there is Applied Materials, Lam Research and KLA-Tencor. The US-based capital equipment builders do manufacture in Asia and Israel, however the bulk of the know-how on how to develop tools to make those chips resides in the U.S.
Until very recently, the U.S. was really underfunded in terms of the number of fabs (microchip fabrication plants) that were built here. There’s just not a huge amount of capacity in the U.S. We have Intel fabs in Oregon and Arizona, Samsung in Texas, Global Foundries on the east coast, and Micron in Idaho and Utah, plus a few others. Now everybody’s realizing that this is a strategic supply, so the U.S. government is advancing a proposal for a $50 billion investment in support of US-based manufacturing and research. For comparison, the Korean government recently committed $450 billion in an effort to make South Korea the world’s biggest chip making base.
Q: How did that split happen in first place?
A: I think it was probably due to cost. Electronics production went from U.S. to Japan, then from Japan to other places across Asia (like Korea and Taiwan) all because production was less expensive. More recently, within the last decade, China has had a big initiative to bring a lot of semiconductor production into their country. They spent a couple of hundred billion dollars to make that happen.
We kind of woke up late in the U.S., but at least there was recognition that we need to have a lot more of those facilities here. This year alone we had announcements from major players investing tens of billions in expanding their production capacity in the U.S. Our share right now isn’t huge, but we’re hoping for it to grow significantly in the next five years.
Q: What is the state of the supply chain in 2021?
A: Unprecedented demand. The recession in 2008 and 2009 put a hold on many things. When the rebound came, people were caught off guard and the upswing was very challenging to keep up with. Compared to that, what we’re looking at here is a totally different order of magnitude.
Chip shortages are everywhere, everybody is struggling and prices are going up. If you want that batch of chips you better buy a lot more than you normally would and pay many times more while you’re at it, or somebody else will pay the price and get the chips.
Of course, at HEIDENHAIN we’re in the same fight because we’re supplying equipment into machines that will build those next-generation chips. Luckily, we are able to keep up, but it’s kind of a cycle: We need more chips, we need more equipment to build the chips, and we don’t have enough chips to build that equipment. That’s kind of where we are.
As for the impact of COVID, one theory is that current demand is partly the result of people not buying in the middle of the pandemic. Maybe this is true to some extent but if you look at the electronics industry specifically, 2020 was solid despite COVID.
Factories did close, a lot of people left their jobs, and a lot of them didn’t come back so now some suppliers can’t run at the same capacity they used to. People notice shortages and everybody gets in line to buy extra. Buyers now are even more motivated to stock up and start hoarding. The combination of demand and people buying things well ahead of what they would normally will suck oxygen out of 2023 I think.
Q: What makes it so difficult to ramp up a facility?
A: On one hand we are creating individual transistors which are the basic building block of electronic circuits, about five billion in a nicer computer chip. As a size comparison, the width of your hair is 10,000x bigger. To make things even more challenging, you are building those layer upon layer and relying on each to be aligned perfectly with the one below it on the order of a couple of nanometers.
To give you an idea, there are five silicon atoms per nanometer so you’re aligning dozens of silicon atoms on top of dozens of silicon atoms. If you’re off by just a few atoms the thing doesn’t work. If you have five billion transistors and 99.99 percent of them work, you wouldn’t have a working computer chip. You have to have tremendous yield and quality control for such a mass of tiny electronic components and miles of interconnecting nanowires.
There’s a reason there are only four or five top-tier suppliers, because the know-how is very specialized and takes decades to develop.
We do see a lot of spending to build up the capacity, but that’s going to take 18 months to a couple years. A build cycle for these tools is typically more than six months. Think about all the different supply chains involved, and then the facility itself which needs to be populated, tested, and validated all before production starts. Because the demand is so tremendous, those companies will be relying more and more on public & private partnerships for incentives or funding to make these facilities a reality.
Q: Where does HEIDENHAIN fit in the semiconductor supply chain?
A: We have a longstanding relationship with many of the biggest companies that produce capital equipment, including all four tier-one capital equipment suppliers who account for more than 60 percent of all machines sold to the semiconductor industry. We also work closely with quite a few of the tier two and three suppliers.
On one hand, our standard portfolio of products offer solutions that others don’t. For some things we’re the only place where you can get the level of accuracy and precision required continually to get those smaller and smaller structures for the semiconductor process to develop. On the other hand, because we’ve been in the business of precision for so long we have a tremendous amount of internal know-how on all aspects related to measuring things on that scale.
Companies we work with shouldn’t just come expecting for us to walk them through our catalog and point to what looks interesting. What we’re actually more interested in is hearing about their highest value problems. What are their challenges? What is it they’re struggling with solving? From there we see how we can leverage our internal knowledge base and fundamental research capabilities to help find solutions.
For example, when we engage with forward-looking units, which are parts of companies that are thinking about future endeavors and how to leverage their existing technology, this is what we talk about. We talk about creating, or making available to them, technology that enables them to go well beyond what they’re thinking today.
Q: How do you work with customers in the electronics industry?
A: We pay very close attention to what’s happening. What are the key industry drivers? We try to gather a very significant amount of application knowledge, not just whether they need five nanometers or one nanometer. What are customers trying to do with their process? What‘s keeping them from getting to the next level? We take those questions and then we look at our products and solutions to see how we can help.
Learn more about our specialized product offerings for the semiconductor industry and unmatched expertise. Our semiconductor engineers look forward to the chance to discuss your unique design challenges. Secure an online consultation spot, we’d be happy to try and help.