A move by Foxconn to expand into the chip business is exactly what the slowing behemoth needs. But there are some big "ifs."
"We want to get into semiconductor design and production," Foxconn Chairman Terry Gou told Shenzhen Satellite TV last week, according to Nikkei Asian Review.
This would be positive for the Taiwanese company if it designed the right types of chips, and if it minimized its financial exposure to semiconductor production. Last week I wrote about the risks faced by TSMC in continuing to spend billions of dollars to chase the bleeding edge of chip technology in the face of a waning end market.
Foxconn's desire to get into chip manufacturing is consistent with its purchase of Japan's Sharp, and the merger seven years ago with two Taiwan display makers. It would boost the company's share of bill of materials -- the components that go into an electronic device -- and reduce reliance on the mere assembly for which it's best known.
Currently, Foxconn has exposure to almost every part of a device's parts list except chips.
The choice of chip type is crucial. There's no way that Foxconn can compete with Qualcomm in communications chips, nor could it take on Intel or Samsung in processors. However, there's a huge market for the little-known yet high-demand chips called application-specific integrated circuits, or ASICs, and application-specific standard products, or ASSPs.
Nikkei noted that Foxconn has been trying to develop ASICs, yet ASSPs would also be on the menu, and here's the difference: An ASIC is designed for a specific purpose, and for a specific customer. Apple, for example, might want a chip that controls the home button on its iPhones. Since that's unique to the iPhone, it might be an ASIC. An ASSP is made for a dedicated function, but one that's common to different devices or customers, such as the chip that operates the USB port on your phone or computer.
To expand in this sector, the maker of iPhones and Xiaomi handsets wants to join with ARM Holdings of the U.K. to create a chip design center in Shenzhen, Nikkei reported. ARM, now owned by SoftBank, designs the core of many of these types of chips, and has created an ecosystem that dominates consumer electronics.
A tie-up would help Foxconn develop ASICs and ASSPs for the Internet of Things devices -- such as smartwatches, connected fridges and cars -- that will be the next growth driver for the industry.
More than any other product segment, IoT is likely to be fragmented and its chips task-specific. If Foxconn can make it easier for customers to develop products by offering the right component for the job, it's likely to land not only chip orders but also contracts for assembly and other parts.
The other "if" is on Foxconn's strategy. If the company sticks with design and leaves the manufacturing to others, like TSMC, then it could do very well. However, Gou may be tempted to try his hand at chipmaking at the leading edge, which would lead to a huge cash burn, weighing on earnings just when sales are slowing. TSMC, for example, expects to spend $9.5 billion on equipment and facilities this year, and last year had $7 billion in depreciation expenses. Foxconn's Hon Hai unit had net income last year of $4.6 billion.
On the other hand, Foxconn may choose to manufacture with older, less expensive technology. Not every chip needs the latest processes -- many ASICs and ASSPs can be produced on relatively cheap equipment. And, assuming good execution, having your own factories can speed up the development and production cycle, especially if quantities are smaller or tailored to specific uses.
Getting into the chip business would close the loop on Foxconn's grand plan to be in every segment of electronics components, and it may just be able to boost that slowing sales growth. If it takes the right approach.
This column does not necessarily reflect the opinion of Bloomberg LP and its owners.
To contact the author of this story:
Tim Culpan in Taipei at email@example.com
To contact the editor responsible for this story:
Paul Sillitoe at firstname.lastname@example.org