The Price Of Consolidation

Rising costs and the difficulty of shrinking features are changing the dynamics of who buys IP, tools and chips.


Consolidation is causing far-reaching changes across the global semiconductor ecosystem due to the size of companies being bought and the dearth of startups to replenish those being acquired.

Coupled with the rising cost and difficulty of shrinking features down to advanced process nodes—many argue that is the largest driver of consolidation—the market dynamics for who’s buying IP, EDA tools and chips has changed for the foreseeable future. That isn’t bad news for everyone. There are winners and losers in these kinds of market disruptions and portfolio realignments, and some companies will see very little direct impact to their businesses. But all agree that when the dust settles, the industry will look far different than before this latest round of acquisitions began.

Semiconductor Engineering has spoken with dozens of executives, analysts and engineers about this subject over the past several months. Those discussions point to a number of important trends:

• Cheap capital due to low interest rates and relatively low valuations for chip, tools and IP companies has made it possible to buy larger companies than ever before, as evidenced by Avago’s proposed purchase of Broadcom. (Avago’s 2014 revenue was $6.57 billion, compared with Broadcom’s $8.56 billion.) At some point interest rates will rise, credit will tighten, and this window will close, leaving combined companies with significantly different operations against which their execution will be measured.
• The rate of consolidation has accelerated throughout 2015. There are typically about 25 to 30 acquisitions a year. In 2015, there will be roughly twice that number. Even more important than the quantity of deals, the size and value of those acquisitions is roughly 10 times larger than in other years, according to statistics compiled by Mentor Graphics.
• While the Internet of Things term has become an overhyped buzzword, evoking images “smart” toothbrushes and connected toasters, the interconnection of technology across markets and within markets is widely recognized as a real opportunity on many levels. Companies continue to position themselves for where they can play in these markets, whether it’s industrial, automotive, medical, home, or others. All of these markets share one common trait—they’re all heavily reliant on semiconductor technology, and companies that are best prepared and first to market will win significant share, even if it’s only within those vertical slices. Acquisitions are being viewed as a quick way to prepare for those markets.

But how long will this consolidation continue? And more to the point, what will be the likely fallout if this trend continues?

“Whenever the level of liquidity is attractive, almost everything is accretive,” said Wally Rhines, chairman and CEO of Mentor Graphics. “The last big boom in liquidity was in 2007 with the leveraged buy outs, and in 1999 to 2000 with the dot-coms. It happens whenever liquidity is available to borrow at favorable interest rates. But you can’t have it go on indefinitely. At some point, this will be self-healing and the market will set the cost of money. In the meantime, the level of acquisitions will increase.”

What’s particularly striking about this round of consolidation is that it reverses a longstanding trend in semiconductors.

“It’s not normal for the semiconductor industry to consolidate,” said Rhines. “The past 50 years has been spent in deconsolidation. Market share among the top 10 is about the same as it was 40 years ago. But if we continue on this consolidation trend, the combined market share of the top 10 will increase 3%. That may not sound like much, but it’s a big number. It hasn’t been this high since 1984.”

Most outside observers, though, see change as an integral part of the semiconductor landscape in particular, and tech in general. Companies working in the field see the changes around them as disruptive, but over time and from a distance, the market is almost always in a state of churn.

“While cheap capital is certainly a contributing factor, acquisitions in the tech sector have built upon themselves and companies have made investments for both near-term acceleration of market positioning, as well as long-term strategic bets to drive overall shareholder value,” said Tom Erginsoy, director of PwC’s Deals Practice. “The semiconductor sector specifically has undergone lots of consolidation over the past several years, which is even more relevant given it is a unique tech sector in which scale is perhaps most critical. The level of deal activity tends to be more volatile than other sectors within technology, but overall we would expect to see the level of deal activity to largely remain unchanged as corporates shed non-core assets in the portfolio accumulated during the consolidation phase. In the long term, we wouldn’t really expect a significant change.”

One immediate fear is that overall R&D spending will be cut as a result of these acquisitions, which would impact investments in new technologies at a time when Moore’s Law is getting harder to follow and the IoT is just getting started. But the results aren’t so black and white there, either. Merged companies do look for ways to trim costs and leverage synergies, but they also manage their acquisitions and sales like any good stock portfolio manager. Avago sold off two divisions to Seagate and one to Intel prior to its proposed to purchase of Broadcom, and NXP has offered to sell off its RF power business prior to last week’s approval by the European Union of its merger with Freescale.

EDA and IP have reflected these macro changes in the semiconductor industry because both of those are required for progress at advanced and established process nodes. Synopsys alone has made 71 acquisitions since the company was founded in 1986. During that time, it has kept its R&D budget at about a third of total revenue—roughly keeping pace with the added revenue from those acquisitions.

“In all fields you go through big variations—we refer to them as bubbles—where buying a pet rock company is really expensive when everyone wants a pet rock,” said Aart de Geus, Synopsys’ chairman and co-CEO “Three months later that company is really cheap. VCs always complain about bubbles and they just hope they’re in one of them. It’s the hope they hit some big return. But in aggregate, markets provide returns to whoever put in the money, so it balances out.”

The size of the deals also raises the stakes for invested capital, though, particularly in fast-moving markets such as the IoT. That can have a direct bearing on EDA and IP companies.

“You have to make sure you’re in the right place to win,” said Lip-Bu Tan, president and CEO of Cadence. ” Everybody is worried about growth because of the challenges they need to overcome to grow. When you get to 10nm and 7nm and 5nm, those are very expensive nodes. So everyone is looking at how to extend their platform. They’re worried about how to extend the area that they’re in—but not so far that they get into trouble. There also is a concern that the target company has the right culture so it can be integrated in the right location, and that the talent that can be brought on board. You try to obtain efficiency in G&A and market reach and focus. Those are issues you need to address in a maturing market. At the same time, there is a lot of real opportunity in the IoT and automotive. So if you’re looking at these markets, you have to question how you can expand your team.”

Who wins, who loses
Behind the headlines, the rate of growth in the semiconductor industry remains in the single digits. By acquiring companies, operating costs can be chopped and the combined company, at least on paper, will have a deeper and longer market reach and more resources to build complex chips at the most advanced nodes.

“As large companies consolidate and then have Op-Ex capacity to buy finFET-class tools, it enables finFET-class designs,” said Jack Harding, president and CEO of eSilicon. “So if you look at Avago’s acquisition of Emulex and PLX, those companies would not be able to afford finFET-class tools. There’s a second-order effect here, which is that it puts EDA tools into the hands of a bunch of guys who couldn’t otherwise afford them. But it also raises an issue for EDA companies. Are they comfortable with the amount of money they’re getting for tools at the leading edge now and walk away from the sub-$1 billion companies, or do they find new business models to serve those smaller companies?”

Most industry executives are in agreement that the outcome of this consolidation isn’t obvious. It will take time to play out.

Charlie Janac, chairman and CEO of Arteris, views the current round of consolidation as a form of “asset rotation.”

“People are trying to strengthen their positions in different segments they feel they need to strengthen,” said Janac. “If they mess it up, they get to sell. The volumes are the same. And if you have a royalty model, nothing changes. The number of licenses or projects may or may not shrink. So IP companies are neutral. With all the M&A transactions, the volumes will stay the same or grow. It’s harder for EDA companies, though. That’s why you see Synopsys putting their chips in IP and software.”

Cadence has made similar bets in IP; Mentor Graphics has diversified into everything from software to wiring harnesses; and Apache has joined forces with Ansys on the system design and simulation side. The challenge is for midsize EDA companies without a specialty in one of the growth markets. “The oligopoly of EDA is not affected,” he noted. “But Atrenta was the last midsize digital EDA company.”

Nevertheless, the customer base has been shifting significantly even if the market for IP has remained stable.

“The expense of putting a chip together is going up, so the number of customers is getting squeezed on both ends,” said Alan Rogers, president and CTO of Analog Bits. “But it’s also not limiting our business. There is more upfront work that’s required now. That’s not consolidation, but there are common drivers between what customers are looking for and what’s causing consolidation.”

He’s not alone in that observation—or in the survival strategy for companies. “Design starts have been going south since 1995,” said Alan Aronoff, director of strategic business at Imagination Technologies. ” The challenge for IP companies is to add specific value. With IP 1.0, everything was about a computational platform, whether it’s a CPU or a GPU. The landscape settled out since then, and it has created new opportunities with algorithms. But algorithms and how they’re implemented are two different things.”

Aronoff said the key is innovation, but he also stressed that not all innovation is profitable. The challenge is identifying the right markets and where innovation will have perceived value within those target markets. Low power is one of the key areas, but that isn’t just about hardware or software. In some cases, it’s how the two work together or how IP algorithms can take advantage of that hardware.

In addition, it’s easy to generalize about consolidation across the semiconductor industry because of the enormous deals that are taking place, but the entire ecosystem isn’t moving in the same direction—or at least in the same directions as they historically have gone.

“Consolidation within the IC capital equipment sector is less frantic,” said David Lam, chairman of Multibeam. “In fact, it’s cooled off a bit since the Applied Materials-Tokyo Electron mega-merger was terminated. But major firms in this space have not relented in their search for small ventures to diversify into new markets or expand their core markets while protecting the existing customer base. The pickings have become increasingly limited in recent years due to the dearth of venture capital. This situation is unlikely to change anytime soon, leading some to predict that major capital equipment suppliers will more likely find opportunity to fuel their growth by acquiring smaller innovative ventures offering competitive complementary solutions that can be potentially disruptive.”

Where startups are being created
Even though it may look as if there are no new startups being created, that’s certainly not the case. VC and corporate investment in startups is significant on a global scale, but it also is being targeted at markets where there is rapid payback.

“Similar to technology as an industry overall, consolidation will undoubtedly change the competitive landscape for a period of time, but innovation and disruption are inevitable,” said PwC’s Erginsoy. “It’s the nature of the industry. New companies with new technologies will inevitably emerge, challenge the incumbents, and become acquisition targets themselves.”

When those startups begin to emerge is uncertain because it’s hard to know what’s been funded until startups emerge either from incubators or from stealth mode. What’s clear is that there is a gap between the last round of acquisitions and the next round of startups, and that the next round of startups will likely emerge with different business plans.

“Startups for bulk CMOS chips have dropped precipitously,” said eSilicon’s Harding. “There is a dearth of investment capital there. But there are startups that are aligning with the IoT. With 55nm and 65nm parts, you don’t have to invest a fortune. So startups can’t get $100 million to build a new network processor, but they can get $5 million to bring an IoT solution to market in 24 months. Low-cost, low-power chips are encouraging entrepreneurs. Whether these are coming from Asia or North America or Europe we can’t tell yet. There’s also a burgeoning IoT chip market in academia, where there’s no lack of IP or smart people.”

Whether those chips are IoT, or simply connected devices within vertical matters is a matter of debate. But there is no shortage of opportunities being created in these markets, and venture funding is active on multiple continents.

“There will be billions of dollars invested in autonomous cars,” said Arteris’ Janac. “And that’s not just for cars. It’s at the truck level, logistics, and for electronic highways. We’re also going to see a big push from rotary to solid-state storage. SSDs are taking over. And there’s also going to be a big push for wearables becoming medical diagnostics, whether it’s for monitoring diabetes, predicting heart attacks or measuring blood pressure. This isn’t just about exercise monitors. What could change the health care system is health monitoring that’s connected to medical suppliers. We’re in a transition now. There are pitfalls and opportunities in any transition.”