Cadence Design Systems Aims to Cash in on New Custom-Chip Era

Cadence Design Systems Aims to Cash in on New Custom-Chip Era
The logo of Cadence Design Systems outside the company's offices in San Jose, Calif., on Jan. 31, 2020. /Stephen Nellis/Reuters
Reuters
Updated:

Microchip design software maker Cadence Design Systems Inc is betting on growth from automakers and other chip users strapped by global supply shortages who face mounting competition from rivals such as Tesla Inc and Apple Inc that design their own chips.

Cadence and rivals Synopsys Inc and Siemens EDA are at the center of a microchip industry shift as cloud computing providers, software makers and others who traditionally have bought semiconductors from a few big companies now want to draw up their chips own in-house.

Tesla, Apple, and Alphabet Inc’s Google are among the leaders of in-house design. Executives across industries have taken note of how custom chips help set products apart, said Anirudh Devgan, who became Cadence’s chief executive last month. The company counts Tesla as a client and analysts say Apple is as well.

Cadence shares fell 5 percent on Tuesday in a down market.

Developing a chip costs around $100 million, but artificial intelligence is reducing costs, even as traditional semiconductor firms keep raising prices, with many chips selling for more than $100 each.

“How many car companies have more than 1 million units? A lot of them,” Devgan said in his first interview as CEO. “At some volume, it’s a no-brainer to do it because of cost, because of schedules, and more importantly, for customization.”

Bottlenecks in the global semiconductor supply chain that have hobbled production at most major automakers for over a year also are forcing companies like Ford Motor Co and General Motors Co to rethink their approach to chip procurement.

The industry is consolidating hundreds of small micro-controller chips that crept into cars piecemeal over decades into a smaller number of more powerful and costlier chips.

Tesla has always used a consolidated approach, and the results contrast sharply with those of other automakers. Despite a global chip shortage, Tesla reported record fourth-quarter production, in part because close control of its chip and system designs allowed the company’s engineers to quickly rewrite code to use chips that were available.

Cadence makes electronic design automation (EDA) software that translates ideas on how a chip should work into the physical layout of tens of billions of transistors crammed onto a few millimeters of silicon. The resulting chips are often manufactured by third parties like Taiwan Semiconductor Manufacturing Co.

In decades past, most of Cadence’s customers were traditional semiconductor firms. But the newer breed of “systems” customers, which dream up full products in which chips play a central role, now account for about 40 percent of Cadence’s revenue, said Jay Vleeschhouwer, head of software research for Griffin Securities.

Cadence has branched out to offer those customers software that goes beyond chip design to helping fit their custom chip into a full product. Cadence has acquired apps for tasks like packaging finished chips to put onto circuit boards and making sure the chips will not overheat and melt in daily use.

The allure of such functions “extends to automotive, aerospace, industrial equipment, all of the makers of products” in which chips eventually reside, said Joe Vruwink, analyst with Robert W. Baird & Co.

With the transformation of cars into rolling computers all but assured, automakers and other chip-design newcomers will face stiff competition when hiring chip architects, who are some of the most fought-over talent in the tech industry.

As little as four years ago, Vleeschhouwer said, EDA companies were not excited about the automotive market because it used less complex chips.

Now, multiple EDA firms “have made pilgrimages to Detroit and other centers of automotive development, and they’re making investments in those areas,” he said. “It’s a consequence of the fact that the complexity of automotive systems, at the system level and the chip level, has become highly motivating.”

By Stephen Nellis