Intermingling Semiconductor and PCB Machine Capabilities

Estimated reading time: 10 minutes

In the midst of electronica, which included the co-hosted SEMICON show, Nolan Johnson spoke with Koh Young’s Harald Eppinger about the convergence of capabilities in the IC and PCB lines at Koh Young. In addition, Eppinger shed light on both the similarities shared between the global regions of production as well as the constraints that each region’s unique requirements place on the manufacturing process.

Nolan Johnson: While Koh Young is known for working in AOI, the company has recently branched out into other new emerging markets like advanced packaging. Talk to us about that.

Harald Eppinger: Correct. There’s a lot of experience based on the PCB and SPI previously for post-reflow. When we look at the last two or three years, we see a huge upward trend in the miniaturization of PCBs. They’ve become smaller, smaller, smaller. And all this development is happening during a trade war and the many other situations currently impacting the economy. For us, there is a component shortage. We’re discussing all the supply chain issues and how big of an impact they have on the local market, and for markets in the Americas, in Europe, and Asia as well. These two trends are interesting to watch. The PCB gets smaller, and the semiconductor enters the world with a leg up over PCBs in terms of complexity, and now the two different worlds are moving closer together. I won’t say they have merged, but even here at the SEMICON show, you can see them blending into the world of integrated circuitry, which has much more functionalities.

If you look at high-tech electronics, we have three common modules: communication, CPU, and RAM. That’s it. Before, we used all the different components to create the functionality on the PCB. Now, a big part is integrated and is delivered either as system-in-packaging or advanced packaging. That’s why when you magnify and zoom into a semiconductor today, it looks like a PCB. On the one hand, that’s the challenge; on the other hand, this is a good thing, because we’ve all learned how to deal with a PCB. We have the experience. The big worry was, can we do these small things? Yes, we can do it. But it’s a matter of getting the passion into this business to ask ourselves why we should leave this business to other regions. Are we capable? Yes. Are we afraid? Yes. Do we have the money to invest in this business? No. We did, but we put it in other regions.

Coming back to the original point, we now have the capability to pursue this in Europe and the U.S. because big sites are being built up in the U.S. We have the CHIPS Act in the U.S. and the Chips Act in Europe, and suddenly a lot of money is around, and this enables us to get back in the game and establish independence.

Johnson: When you say it puts us back in the game, are you referring to regions, such as Europe and the U.S.?

Eppinger: Yes. Everybody involved in electronics is now able to reconsider the home game. I have attended several events and discussion these last few days, and they all say their labor costs and the cost for production are the same everywhere. There was also a big CEO at the table who commented that if the conventions in the different regions are different you cannot be competitive. If you go outside Europe and you get a huge space free of charge, you get a tax-free commitment for three months, or three years, this cannot be compensated by the investment. Daily costs—those you can compensate.

Thinking about that dynamic, if the balance is equal, we can and will be 100% competitive. But if a region can’t commit to that, you get discussions like the current one about the European Union and why it’s concerned about the different ways of putting government money into this business. If America is more interesting, with higher rates, then it’s logical for companies to move to the U.S. If it’s more attractive in Europe, they will put the money in Europe. But this is just about investments in machinery. In other regions outside Europe, there is maybe free land, or low-cost rents, and those are the costs we have to afford. Everybody now has the chance to position themselves.

We see a lot of activities in SEMICON just around the corner, in the other exhibit hall. It’s amazing. We are a main sponsor there. We gave a speech yesterday about the capability of our new machines, and we saw a strong interest. The quality and the next level of wafer, the next level of density, is forcing even higher automation and less defects. Those defects are always costly, but on the wafer level, on advanced packaging, it’s even more. When we talk about the PCB, it’s 20,000 to 30,000 components or dots, and suddenly you have two million dots in a much smaller package. If you look at this and think I have a yield of 99%, that’s a disaster in semiconductor. That’s why a .1-.2% improvement is a significant win and return on investment when you implement the right tools.

Johnson: With all these dynamics, how do you steer the research and development work at Koh Young to respond to these needs?

Eppinger: The good thing is we do not need to respond now because this business has already been ongoing for years at Koh Young. We have been doing business with the big CPU manufacturers in Southeast Asia and other regions for over two years. There was no business in Europe because the semiconductor business moved to Asia. That does not mean that Koh Young was unprepared, however; because we are an Asian company, we are close to that region and are familiar with the dynamics regarding semiconductors in that area. All that was already in place two years ago. The machine is now released to the European and to the joint American/Mexican market. This is good for us, as we don’t have to put too much into R&D on that now. Now, we can put the R&D into the next generation, specifically into process improvement and KSMART, where we generate this digital twin of any application. That’s empowering.

Johnson: So even more of the factory of the future—or digital factory—capability goes into the tools. Is it fair for me to assume that, with this sort of equipment for advanced packaging already in semiconductors, and already in Asia, that current equipment is optimized toward production rather than a more high-mix low-volume approach? Does the equipment need to change to adjust to the new territories?

Eppinger: That’s a very good question, and to answer it, I would like to take an example from our regular business. We have the big guys on board, the mass production for iPhones, everything. We have been talking about mass production 24/7, but the very interesting part is even these companies have to go for a high-mix environment due to the component shortage. Suddenly, they need very high flexibility because they can’t get the components. They had to have initially one or two changes a month, and now they have two to three changes a day. That’s equivalent to the need of a mid-size account. It’s 100% comparable, and the high volume made it easier to ramp up process improvement because we can focus on something. The KSMART tools we implemented then  collect all the data, which allows us to visualize the process. We can see every single process in the production chain, which enables us to know exactly how good we can be.

In high volume, we just would like to see a very stable, flat line, so you want to remove the bandwidth of the tolerances and production deviation by continuously improving the environment. The high-mix environment has had to do it daily for years. I would say this is common in America and Europe because we don’t have mass production; all that has moved to low labor cost areas, but on the other hand, we have the flexibility. That’s the power mid-size accounts have over the super big accounts. They are the basis for healthy economies. They are not reacting immediately to a crisis. They are privately owned. They make wise decisions. They are not always just profit driven. They trust in a world where they can reach their goals and position with good quality, with innovation to the market. The combination of this initially high volume is enabling flexibility, which is now also seen as a very useful thing by the high-volume guys. This is a daily business, especially in Germany where you’ve got one to two line companies. We’re not talking five to 10 lines.

This flexibility is not just about the machine. Thinking “I would like to catch the defect” is the wrong approach. When I said the miniaturization was driving our business, I mean that 10 to 15 years ago we could take something in our hands, but today we can’t even see it. That means there is no chance for repair. There is no chance for rework, so catching a defect is the goal of our equipment. It’s much more important to understand exactly when the deviation is starting so you can fix the mistake or defect by tomorrow.

Johnson: AOI is no longer about pass/fail. Now, it’s also used to modify the process so that you’re back inside the parameters of the process window.

Eppinger: Exactly. This is a kind of closed loop, with preventative actions, preventive maintenance, control, and even knowing much more about how to utilize the equipment to the etch. We cannot buy a new system, equipment, or factory every year. But even older generations of equipment can deliver. If you regularly perform maintenance on your car, it can last you over 100,000 miles. If you think you can save money by not changing the oil on time, well, your car may temporarily get you from A to B, but you might have to consider how long that will last. I think this is something we have to understand: it’s a big transition from the gatekeeper mindset—“let’s catch the defect”—into a more proactive, full understanding. Just yesterday we had a forum discussion at the show on how to close the skill gap in electronics. We have experienced people here with many years of experience in the business. We need new people onboard, but we can no longer afford to let them learn over years and gain that experience learning hands-on. That is no longer possible; we must bring the young people on quickly in operation, and these kinds of tools are helping them understand the process and how to do things right.

A senior guy looks at the line and says, “Move this here. Pull that. Press this button. Fixed.” But that’s intuitive; it might not be based on facts. It might be also potentially the wrong decision—maybe it fixed the problem today, but it may have created a new problem for tomorrow. Playing with this visualization of the data and processes enables more people to take the right action which, at the end of the day, leads to a gain in profitability. Less hands-on means fewer people on the shop floor. Fewer people on the shop floor also means fewer potential human defects on the shop floor, so that enables us to be competitive. If you look at the big sites, they’re fully automated, and you can do full automation in Europe, America, and Asia. Some of the hard workload in some applications need no labor costs. In this puzzle, there’s a lot of small pieces we must put together, and when it all fits together, it enables competitiveness. The competitiveness within our customer portfolio is the most important thing, from small accounts to mid-sized accounts. This applies to applications in medical, military, aerospace, and industrial automation.

We all know automation is the future and that it requires high dense electronics; at the same time, we need a huge power supply. Before it was sensitive electronics, dense electronics, and power industrial automation. It was like two worlds. Now we have a high-end microprocessor controlling the whole charge, temperature, and energy supply; at the same moment, when somebody says, “Oh it’s a free highway here in Europe. I need to test the car,” we have hundreds of amperes. This new mix of technology is challenging for us as a vendor, but it also enables a future for our customers.

Johnson:  Harald, very insightful. Thank you.

Eppinger: My pleasure.