Acquisition sees boom time for semiconductor equipment

Interviews |
By Nick Flaherty

Three years ago, US-Israeli printed circuit board inspection company Orbotech bought a semiconductor equipment maker in Wales in a surprise move. Today, SPTS retains its branding and is seeing a resurgence in LED manufacturing equipment and a range of opportunities for a new plasma dicing technology.

“In general we must still be in the honeymoon period,” said Kevin Crofton, whoo was the chief operating officer at the time of the acquisition and is now president of SPTS and a vice president of Orbotech. “Usually the first thing a US company wants to do is overlay their business processes on you and make you their own . When Orbotech bought us they were very clear they wanted to have an arms length relationship for the foreseeable future and learn from what you do, and they have done exactly that. We are still a stand alone business with our own front and back office so I’m beginning to believe its an Israeli trait and not just an Orbotech trait. Their first step is to understand the business and if it is running well don’t break it.”

There’s three key things that the acquisition has brought, he says. “Firstly, we are now part of a billion dollar company and size does make a difference when you are dealing with the large semiconductor device manufacturers, you automatically get across the starting line whereas when you are smaller there are some companies that don’t give you the time of day unless you have remarkable technology.

The second benefit has been more resources for technology development.

“When you as a $300m company you can take one big bet every couple of years,” said Crofton. “With Orbotech we are able to take 3 or 4 big bets every two years or so, so we have flexibility to try things that maybe we wouldn’t have done in previous years, and this allows us to expand into new markets faster.”

“For example we are developing a plating system that uses electrochemical deposition (ECD) for the advanced packaging arena. That’s a wet chemical line, which isn’t something we have done so far, but it’s a natural extension of the processes that we run, create a via, etch the via, putting down a PVD film. The next step is then ECD, so if we can have in integrated process with a new tool. That opens up a $200m to $400m market for us, so it’s a big bet but with a big potential reward.”

Technology sharing

Plasma dicing
Plasma dicing has been a key new area for SPTS

There have also been advantages of sharing technology with Orbotech, from image processing to 3D printing.  

“As Orbotech is a PCB inspection and repair company they have some core competences that might be usable in our space to solve some problems,” he said.

The first of these is using their PCB direct imaging technology for wafer formats.

“We are also using a 3D additive manufacturing process to create underfill dams on a die that needs to be supported,” he said. “With 3D printing we can also put a an isolation layer selectively between the balls on a BGA substrate and both of those solve fundmental problems that our customers have making their devices. Orbotech make the print heads with the UV curing activity – that’s where the real science comes from for device marking and many other areas. Those are big wins from the merger activity.”

Another big move has been to take the existing plasma vapor deposition (PVD) technology in to new areas such as into advanced packaging for fan out wafer packaging. “We are by far the market leader in PVD for fanout wafer level packaging, and we are the process tool of record for PVD at the world’s largest foundry as well as 6 out of the top 7 packaging companies,” said Crofton.

“Now we have also launched a plasma dicing capability for both carried wafers and normal wafer and also no tape frame assemblies,” he said.

This is particularly important for small die for RF and LED makers with thousands of devices in a wafer. Using cleaner plasma cutting can significantly improve the number of devices on a wafer and the yield.

“When you have dicing lanes of the same dimensions as the device, if you can take that dicing lane down to 10um you are using less die area. You can’t do that with a blade or a laser because of the thermal effects so you get more useable die per wafer. One high brightness LED company has seen an increase in usable die of 30% as a result, which is huge.”

This is also helpful with thinner wafers used by MEMS makers. “When you thin wafers to 50 to 70um thick, they are susceptible to edge fractures, and plasma cutting avoids the cracking. For us it’s the most popular demo machine we have.” The technology is being used by SPTS customer ams in Austria for MEMS microphones developed by Knowles that go into the iPhone 7.

Future growth

A MEMS microphone in the iPhone 7 cut with plasma dicing from a tear down by SystemePlus Consulting.

These new technologies are deliberately targeting changes in the industry in several areas, he says.

For example, the transition to 7nm will also drive fanout adoption as more devices are pad limited and need the fanout from connections across the chip to the package pins.

There’s also some fundamental market drivers for SPTS customers, particularly in automotive. Autonomous vehicles, advanced driver systems, Lidar, radar and all the sub-systems. “Automotive will, l believe, replace the smartphone for driving LED, power and RF development,” he said.

“Then there’s smart lighting that’s driving companies like Osram, Nichia and Lumentum, they are absolutely booming right now, for homes, for automotive and you don’t hear them talking about TV anymore.”

But being small also helps, he says.

“If you think about RF, power, LED and MEMS, those four areas are on 6in wafers or below on substrates such as GaN and SiC, all those things that a foundry like TSMC doesn’t want to deal with but it’s a perfect scenario for SPTS as it fits into our niche markets,” he said. “Those companies are now buying new equipment and we have always made a business of selling new equipment into these market.”

“We sell all the SPTS technologies into all those markets, we share our product and technology roadmaps with them and have a dialogue about where they want to be 2 to 5 years from now for the performance of their devices. For example, in the MEMS space for a while ST would specify a shape and we would work to provide that, and three years later that product has moved on with a refined shape so we will give process engineering support to get to that new shape – the big companies in this space won’t dedicate the resources to make that happen.”

Related stories:


Linked Articles
eeNews Europe