Two huge mergers have shaken the global semiconductor industry in recent weeks: Avago bought Broadcom for $37 billion, and Intel purchased Altera for $16.8 billion. The deals are part of a major transition the semiconductor industry is undergoing, as evidenced by the fact that the Broadcom-Avago merger is the largest ever in the chip sector, while Altera is Intel’s largest acquisition ever.
- Israel's problem is too few Intels, too many Wazes
- Chip maker TowerJazz rises from the ashes
- Panasonic to sell 3 Japan chip plants to Israel's TowerJazz
- Avago buys ex-Israeli company CyOptics for $400 million
The consolidation of the semiconductor industry, which laid the foundation for the entire digital revolution, has very interesting consequences for the Israeli high-tech sector. Almost all the world’s leading semiconductor firms have large research and development centers in Israel, alongside which a large number of local semiconductor companies grew.
Semiconductors are one of the most important areas of specialization for local high-tech firms. In light of the great interest in recent times – both in the media and among Israeli policy makers – concerning the need to encourage the growth of local companies rather than making do with startups that end with impressive exits, the consolidation of the global semiconductor industry will likely make growing local firms an even more difficult proposition.
Mergers in the semiconductor industry are nothing new. Over the past two years, though, their pace has accelerated and the value of companies bought since December 2013 has reached some $80 billion. “The semiconductor industry is maturing. The growth rate of the industry is dropping, the profitability of the companies is falling and the complexity of chips and their manufacturing cost is rising. At the same time, there is very strong pressure to lower the price of chips,” says Shlomo Markel, a vice president at Broadcom, which has extensive R&D operations in Israel.
“The assumption is that the next stage of the digital revolution is the Internet of Things – in other words, connecting objects to the Internet using very cheap components that use very little energy. In order to reduce the price of the components, economies of scale are an asset both in development and manufacturing,” adds Markel.
“Size in the semiconductor industry definitely matters,” confirms David (“Dedi”) Perlmutter, formerly the highest-ranking Israeli at Intel internationally and today the head of the Eucalyptus Growth Capital fund, which is raising $300 million.
Perlmutter says semiconductors today are entire systems on a chip, integrating a great number of types of intellectual property. The goal of the recent mergers and purchases is to create complete product lines that will allow a single huge company to create a critical mass for itself and provide a complete solution for the makers of consumer products, like Samsung or Apple, he says. To continue to grow, semiconductor companies need to increase market share. Small and even mid-sized companies, such as Broadcom, have to sell in order to survive, he adds.
Nonetheless, Perlmutter admits he was surprised the Avago-Broadcom deal. “I thought Broadcom wanted to remain an independent company. I assume the main reason for the merger with Avago was to better compete against Qualcomm in the cellular arena.” But Perlmutter doesn’t believe the number of Broadcom employees in Israel will shrink as a result of the merger.
Knowledge, creativity and risk taking
“The semiconductor industry in Israel is at a very interesting point,” says a senior executive at one of the multinational companies operating in Israel, who asked not to be identified. “We’re good at three things: the creativity of our engineers; taking risks; and professional knowledge. We have no advantages in manufacturing or price. We need to understand what the large semiconductor companies are looking for. They don’t need more marketing or sales people. Everyone is looking for technologies that will give them an advantage over their competitors in contending with the challenges set by Moore’s Law.”
Moore’s Law, which was postulated by Intel cofounder Gordon Moore in 1965, says the number of transistors in a dense integrated circuit has doubled approximately every two years, and will continue to do so for the foreseeable future.
“Israeli engineers are exactly what [the large firms] are looking for. Because of this, Apple came to Israel. And that’s the reason Israel interests the Chinese manufacturers so much. I believe this trend will only strengthen the Israeli ecosystem in the sector,” says the executive.
Despite the praise he heaps on the local industry, the executive warns that the chances of a local semiconductor firm developing into a major player are smaller than ever. Among Israeli semiconductor companies, Mellanox Technologies and EZchip Semiconductor remain independent; Anobit and PrimeSense were sold to Apple in 2011 and 2013, and Annapurna Labs to Amazon in January, to name a few.
“In order to create a new chip, you need to spend at least $50 million and quite a bit of development time,” says the executive. “So, angels and venture capital investors are in no hurry to invest in semiconductor ventures, and the number of startups in the field in Israel and around the world is very small. This doesn’t mean that initiatives in the sector won’t succeed. Most startups will simply be forced to sell out at the development stage, even before they’ve created a working product with an investment of tens of millions,” he predicts.
The right chip for Intel
Intel has had a hard time growing in recent years, after missing the boat in the mobile revolution. As the desktop computer market shrinks, Intel is trying to refocus growth to the data center computing market, where it remains the main supplier of processors for servers. Among its largest customers are Google, Amazon, Facebook and Microsoft, companies that have huge data centers and provide cloud computing services.
In light of the growing competition between its customers, Intel is trying to sell each of its large clients a slightly different product, tailored to the customer’s data centers and its proprietary software. Microsoft could buy a chip set from Intel, including the microprocessor, and customize it later for its computing center and cloud solutions, while trying to create a competitive advantage for itself over Google or Amazon, for example.
Altera has already pioneered this approach, with chips that can be programmed after they’ve been manufactured and designed for smaller production batches and adapted to a customer’s individual needs.
Until it announced its acquisition of Broadcom, Singapore-based Avago operated quietly. It was the product of a number of mergers in the semiconductor industry, starting in 2005 when a group of private investment funds, headed by KKR and Silver Lake Partners, bought Agilent Technologies, the semiconductor division of HP. That same year, the company had its initial public offering on the Nasdaq in New York, and in 2013 bought the semiconductor manufacturer LSI. Over the past five years, Avago shares have climbed almost 600%, making the stock a valuable currency for M&A deals. Over the same period, Broadcom shares rose only 76% until the announcement of the sale to Avago.
In the agreement with Broadcom, Avago will pay $17 billion in cash to Broadcom shareholders and pay another $20 billion for Avago stock. Avago has only $2.5 billion in cash at the moment, so it intends to finance most of the purchase through raising debt. This is characteristic of a private equity fund deal.
Avago develops and sells thousands of types of chips, which are used in smartphones, Internet equipment, equipment for communications network providers and storage. Broadcom is stronger in manufacturing networking equipment, television converters, modems and network equipment for service providers. Sales of the merged company will reach about $13 billion a year, says Perlmutter. The goal of the purchase is to reach critical mass. Each of the companies complements the products of the other, and together they’ll supply customers such as Samsung, HTC and Huawei. Instead of buying knowledge from other companies, the merged firm will be able to offer complex chips based on its own knowledge. Broadcom has been trying for years to enter the mobile market and this sale could help it do so, says Perlmutter.