Interview: Cypress CEO looks for the perfect balance
Cypress is an interesting chip company because at $2.5 billion annual revenues, while not a giant of the semiconductor industry it still offers a broad range of products. It started back in 1982 as a vendor of memories and programmable logic, which it still offers in the form of programmable SoCs (PSoCs). By a series of acquisitions and incubations the company has moved into MCUs, wireless, PMICs, touch sensors, timing solutions, and interface ICs including USB.
It offers many products but often quite application sector specific so we started by asking El-Khoury if Cypress is too broad – or not broad enough.
“Cypress has the balance that I want it to have. The types of products and market sectors are in balance,” El-Khoury shot back. “For automotive: MCU, memories, PMICs, SoCs. We have every one of those. For the world of IoT there needed to be the RF next to it, which is why we bought Broadcom’s IoT business. We are one-third automotive, one-third consumer, 17 percent industrial and 17 percent extra pieces.”
The reference to consumer is also interesting because the volatility of that market and the recent saturation of the smartphone market has had many companies rushing to move their capabilities into steady markets such as automotive and industrial.
“There is a difference between exposure to consumer and exposure to one or two companies,” said El-Khoury. “Cypress is not in that boat. We do not have one 10 percent customer and we have walked away from orders worth $150 million over two years. What we have is ‘sticky’ high-value consumer customers like Netgear and Amazon.”
With a history of acquisition it would be interesting to know what will come next.
Next: You could argue….
“The merger with Spansion in 2015 was big [valued at $5 billion]. But we don’t do acquisitions just for scale. We are very disciplined about what we do. We did acquire the Broadcom’s wireless IoT business in 2016 [valued at $550 million]. We are now in radio because of our MCU business. We also acquired a software company in 2018 [Cirrent Inc.]. We’re always looking. Standing still is not an option.”
El-Khoury will not release any information on any upcoming deals but is willing to set the scene. “Our debt is the lowest it has ever been and with that liquidity we’ve been buying back shares. You could argue that Cypress is ready to do something.”
This leads to the observation that a wave of consolidation that swept the semiconductor industry in 2015 through 2017 seemed to recede in 2018.
“There is always consolidation. In 2016 and 2017 you were reading about it in the public domain. China was attempting to buy a lot of companies but often didn’t close. Now China is not putting players up for obvious reasons,” said El-Khoury.
El-Khoury pointed out that Cypress is in a sort of middle region between small, which he defines as less than $1 billion in annual sales, and big, which he defines as greater than $5 billion in annual sales. But he is sure this will change. “In five years’ time Cypress will be much bigger – either because we have acquired or because we have been acquired.”
“We are growing, stable and predictable which puts us in an area where we can be disciplined. In terms of acquisitions I can wait. I am not going to overpay. But for the right target we can move fast, assimilate quickly and monetize the deal. We can be the consolidator.”
What market areas can Cypress expand into?
“IoT is not so much a market as a capability that runs across all our markets. It is the smart-home, the connected factory, the connected car,” he said re-iterating Cypress’s consumer, industrial and automotive interests.
Next: Don’t get distracted
And on the technology front does Cypress need to develop artificial intelligence and non-volatile memory?
“I don’t follow the next shiny thing,” said El-Khoury. “But I do see the significance of ‘edge intelligence’,” he added. “We need to be adding capabilities at the edge that can run on batteries. As to memory; we are the King of Memories. We are number one in automotive flash.”
There are a number of companies coming through with variants of resistive RAM, magnetic RAM, phase-change memory, to provide non-volatility at geometries that flash is unable to achieve although often in the form of intellectual property as embedded memories, rather than discrete.
“The thing is we don’t dabble; we focus. We can choose whether to develop something organically or license it in. It depends on how fast we want to do it. Machine learning could be an edge device in your smart home. But the use cases are lagging the technology.” So, that suggests there is time to develop something organically? El-Khoury agrees but adds “It’s about return on investment and the right time to market.”
“MRAM is probably the most viable non-volatile memory but does not serve all use cases. It only goes to 85 degrees C and maybe 105 degrees C in a couple of years.” These low operating temperature limits make it unsuitable for many automotive and industrial applications. “Cypress has embedded flash memory. We have SONOS [silicon-oxide-nitride-oxide-silicon] and eCT [embedded charge trap] memory. At 40nm we offer 8Mbytes of embedded flash. At more advanced nodes customers tend to use external flash, El-Khoury said.
There is an argument that customers use external discrete memories because no non-volatile memory has yet emerged with the right balance of speed, density and endurance and applicable at 20nm and below, but that when it does it will sweep the field.
Next: I won’t bet the ranch
There are many next-generation memory technologies that are interesting, said El-Khoury but added: “I won’t bet the ranch. All of these technologies have already been in the works for years.” The implication is Cypress will let others do the heavy lifting and gain access to the most appropriate embedded memory technology when it is available form a foundry.
Cypress runs a mixed manufacturing business model with its own 200mm fab in Austin, Texas. In 2017 the company sold its Bloomington, Minnesota wafer fab, which became Skywater Technology Foundry, with a multi-year supply contract to produce legacy products for Cypress (see SkyWater foundry formed from Cypress fab sale). “All future [manufacturing process] technology at 40nm and below are foundry,” El-Khoury added. El-Khoury said that if embedded MRAM or ReRAM with the right characteristics emerged it would likely be available from a foundry.
The ending of an economic cycle, together with exceptional items such as the US-China trade war, seems to be leading forecasters to the conclusion that 2019 will be a difficult year. El-Khoury doesn’t disagree but said that from Cypress’s point of view it doesn’t change anything. “I don’t make a knee-jerk to short-term events. Things maybe slower in the first half of 2019 and build again in the second half. But all our peers see the same thing. We will weather through it and not change our strategy. We will invest. Small companies can’t afford to, but I will double-down on demand creation; press the accelerator because we have the opportunity.”
How does El-Khoury view China as a market and source of competition?
“China is another one of a list of competitor countries – Japan, South Korea, Taiwan – but leading-edge development still happens in the US. China is still one or two generations behind [in manufacturing process technology]. There’s also a lot of talk about intellectual property as the next frontier and the US has a lot of IP. China needs to invest faster to catch up,” said El-Khoury.
“China is important as a market. We just celebrated being in that market for 20 years. But it is important that we protect IP there just as we do everywhere else. But we welcome competitors on a technology basis, because we will win where we focus.”
Related links and articles: