The White House is a sideshow; Intel must decide what it wants to do; link here.
From the linked article:
Five months in and Lip-Bu Tan is already fighting for his job...
The identity question—whether Intel wants to be a chip designer, a manufacturer, or remain both of those things—was possible to gloss over in Tan’s early days as chief executive officer ...
Now, with tensions bubbling in Washington, D.C., and between Tan and his board, his lack of a long-term strategy has become much harder to ignore.
Tan might ultimately survive the D.C. blitz. But his indecisiveness in his relatively short time in the post has merely highlighted how deep Intel’s problems run. The company was once the undisputed king of advanced chip-making—in both design and manufacturing. But strategic missteps that began at least three CEOs ago have culminated in a company that has lost its manufacturing edge to Taiwan Semiconductor Manufacturing Co., and its chip-design edge to rivals such as Advanced Micro Devices — and even to one-time customers like Apple.
Tan’s predecessor, Pat Gelsinger, was ousted last year after his multiyear turnaround effort sputtered. Unsuccessful though it was, Gelsinger at least had a plan: spend hundreds of billions of dollars to build up Intel’s manufacturing capabilities, get back into the chip-making technology lead and start competing with TSMC. Gelsinger was appointed in February of 2021 and presented a detailed strategy for Intel’s revival a month later.
That plan might have had better chances if market forces didn’t work against Gelsinger—most notably a sharp, industrywide pivot to artificial intelligence computing that left Intel in the lurch. Data center budgets went to Nvidia’s AI chips, and not as much was being spent on Intel’s server central processing units.
The company’s annual revenue plunged by nearly a third over the last four years, while Nvidia’s sales are now double Intel’s at its peak.
All that has left Tan with the unenviable position of fixing problems he didn’t create. But his strategy so far seems little changed from that of his predecessor’s approach of trying to fix everything, all at the same time.
That is a tall order, especially given that Intel is still burning cash while its manufacturing side has lost $13.6 billion in the 12-month period that ended in June.
And the company made clear on its second-quarter call last month that its latest chip manufacturing process called 18A is going to be mainly used for its own internal products—meaning no major external customer has yet signed up to have Intel make its chips.
A logical way forward for Intel might be to break itself up, hiving off its manufacturing operations into a separate company from its chip-design operation. That would follow a longstanding industry trend where companies either specialize in manufacturing or designing chips, but rarely both.
Many of Intel’s recent struggles stem from a lack of boldness in shifting along with the market away from its bread-and-butter PC and server chips, be it into mobile-phone chips or AI chips. But Intel isn’t entirely a stranger to decisive moves: its decision to exit from the computer-memory business in the 1980s helped push it to focus on the CPUs that drove the PC revolution of that decade and beyond.
Tan isn’t a career Intel guy—he is the company’s first CEO who didn’t previously work there.
But both he and Intel’s board need some of the decisiveness of Intel’s storied CEO of the ’80s and ’90s, Andy Grove, who was known for saying “most companies don’t die because they are wrong; most die because they don’t commit themselves."
No comments:
Post a Comment