When Pat Gelsinger returned to Intel as its CEO in February 2021, he took over a company that had been battered by mismanagement and weakened by competition.
Intel had lost significant ground in process-node development to Taiwanese chipmaker TSMC. While TSMC was making transistors at 7nm, Intel was struggling to get 10nm. AMD was besting Intel in both client and server performance and taking more market share with each passing quarter. Nvidia was on its steady march of domination in the GPU market and gaining mindshare as the ultimate AI processing vendor.
Meanwhile, Intel had missed key deadlines, most notably Sapphire Rapids, a major overhaul of its Xeon processor that was supposed to ship in 2021 but only shipped this year. Intel also had lost vital talent, including Renee James, who now heads Arm server-chip vendor Ampere, and Jim Keller, the lauded chip designer who fueled AMD’s comeback with the Zen microarchitecture.
Looking to recapture Intel’s semiconductor dominance, Gelsinger announced plans for ambitious fabrication upgrades and an aggressive timeline for delivering on its chip roadmap.
So, after two years of navigating pandemic surges in demand, supply backlogs, and sluggish demand post-Covid, how is Intel doing? Not bad, say analysts.
What challenges has Intel faced since 2021?
The company benefited from a morale boost when Gelsinger was named CEO, but its other problems—late products, lagging the competition—weren’t going to be fixed overnight.
“He’s put a plan in place that is going to take time, and anybody who was paying attention when this all started should have known that it was going to take time,” says Glenn O’Donnell, vice president and research director with Forrester Research. “The investment that’s necessary to turn Intel around is going to be substantial.”
Adding to the challenge are global market conditions. The tech industry today faces rising inflation, fears of a recession, and a lull in sales after a pandemic-fueled flurry of spending when enterprises shifted to remote working.
“We as an industry are all navigating through one of the most complicated macro environments ever,” Gelsinger said via email. “We are balancing shifts in the world order, shifts in economic and social theory, managing out of a global pandemic, all while also planning for the future.”
“Things such as the deterioration of US-China relations and the war in Ukraine and resulting disruption in EMEA and the energy market are well beyond anything I imagined walking back into Intel’s headquarters. Globally, we continue to see macroeconomic headwinds and economic uncertainty,” Gelsinger said.
Return to engineering roots
Gelsinger was 18 years old when he joined Intel in 1979, and during his 30-year tenure, he made no secret of his ambition to run the company one day. That ended in 2009 in a power struggle that he lost.
Gelsinger left for EMC, where he served as COO before moving over to VMware as CEO. Under his leadership, VMware nearly tripled its annual revenue to $12 billion and executed more than 30 acquisitions. The company expanded from basic hypervisors into networking, cloud, security, containers, and 5G. Glassdoor voted him CEO of the Year in 2019.
When Gelsinger returned to Intel, he assumed the CEO helm from Bob Swan, who joined Intel in 2016 as chief financial officer and assumed the top job in 2019. Swan had stabilized the company; he improved morale somewhat, started selling off non-core businesses, and oversaw a few smart acquisitions, like Barefoot Networks. But he wasn’t a long-term solution.
Gelsinger has the engineering credentials that Swan lacked. Chip firms are often led by engineers, and Gelsinger earned a master’s degree in electrical engineering from Stanford University. AMD CEO Lisa Su holds a doctorate in electrical engineering from MIT, and Nvidia CEO Jensen Huang has a master’s degree in electrical engineering from Stanford.
Gelsinger inherited a significant challenge, but his arrival created some momentum, says Daniel Newman, principal analyst with Futurum Research. “He was clearly well respected in the organization and had a vision. I think he helped to create a bit of confidence, but he had some serious challenges with personnel,” Newman says.
O’Donnell says Gelsinger has done a good job of communicating to the market where Intel aims to go. Execution, however, is still a big question mark. Meanwhile, AMD and Nvidia seem to be executing nearly flawlessly, with few missteps or missed deadlines.
Focus on fabs
Much of Gelsinger’s focus to date has been on fabrication. In March 2021, Intel announced it would invest $20 billion to build two more plants in Chandler, Arizona, where its newest semiconductor fabrication plant, Fab 42, opened in 2020. A year later, Intel announced plans to invest $20 billion to build two new semiconductor manufacturing factories in Ohio.
Intel also plans to take advantage of grants and subsidies from the $52.7 billion CHIPS and Science Act, which is meant to reinvigorate US chipmakers and includes incentives to expand or build semiconductor manufacturing facilities in the US.
On the product front, Gelsinger’s promise in 2021 to release five new process nodes in four years was met with some skepticism, especially after Intel went several years stuck on one process node (14nm). “We took an aggressive goal of delivering five process nodes in four years, and we are on track,” Gelsinger said.
With each generation of process node, transistors get smaller, faster, and more energy efficient. Intel’s 7nm process, Intel 7, is now in high-volume manufacturing for both client and server; Intel 4, the successor to Intel 7, is manufacturing-ready; Intel 3 is on track for production later this year; and Intel 20A and 18A have taped out (where real silicon prototypes are created from the software designs of the chip) with silicon running in the fab on internal test chips and those of a major potential foundry customer, Gelsinger said.
At the same time, Intel is getting back into the business of making chips for non-Intel customers through the Integrated Device Manufacturing plan (IDM 2.0) program.
With the emphasis on domestic chip-making capacity, Intel could score big by having its fabs in the United States, notes Newman. “If we start on-shoring more, this could pay big dividends for Intel. But we need leading edge fabs in the US,” he said.
Gelsinger cut costs by selling off lines of business.
While Gelsinger has been plowing billions into the fab business, he has taken an axe to other areas where Intel is not profitable.
He has shut down several product lines, most notably Optane persistent memory. Intel has also sold off the McAfee security division (although the process began before Gelsinger arrived), the Intel drone business, NAND flash products, RealSense visual sensors, and Intel Sports.
O’Donnell has mixed feelings about whether Intel should have stuck with Optane, but if Gelsinger is trying to refocus the company on core capabilities it maybe necessary. “In the long run, it may hurt them, but in the short run, I think it’s the right thing to do,” he said.
Newman sees that cutting lines of business that aren’t central to Intel is important now. “I didn’t see Optane as a core focus area for the business. So, I agree with the idea of moving away from projects that aren’t core when you’re in a reorganization period,” he said.
Another area where Gelsinger has reprioritized spending is stock buybacks. While a stock buyback might boost a company’s stock price in the short term, it represents money not invested in R&D, for example. From 2019-2020, Intel’s spending on buybacks ($27.8 billion) surpassed its R&D spending ($26.9 billion), and Gelsinger promised to end that. True to his word, Intel has not made any stock buybacks under his tenure.
Wall Street is being patient for now.
Server sales in general are cyclical. Four to six quarters of high sales, usually around the time that new processors are released, are typically followed by four to six quarters of low sales as companies deploy the hardware. Right now, the industry is in the latter stages of the slow cycle. Powerful new server processors on tap from Intel and AMD might help goose server sales this year, but there’s no guarantee.
“The macro-economic situation is probably the hardest thing working against him,” Newman says. “But at the same time, maybe the silver lining of the slowdown is he’ll have everything in order before the next big boom in chip buying cycles.”
Intel’s stock has languished over the last few years, while AMD and Nvidia stock prices have soared. Intel’s revenue for the fourth quarter of 2022 fell 32% year on year, leading to a $664 million net loss for the quarter. On a call with analysts, Gelsinger issued a warning about Intel’s fiscal Q1 2023 results, due April 27: “Our results and our Q1 guidance are below what we expect of ourselves.”
Ben Bajarin, CEO and principal analyst of Creative Strategies, follows the finances of semiconductor firms and says that there has always been a belief among financial analysts that Gelsinger is the man for the job, and they are patient for now.
“I think there was a lot of positivity that Pat was the right guy to right the ship and really start to hit their execution goals. I definitely don’t think they expected this to be an immediate fix,” Bajarin said.
Something that may quiet grumpy investors is the recent spate of news from Intel about forthcoming server and AI products. Intel laid out new products and gave demos that showed it remains on track to keep its four-nodes-in-five-years promise.
“It is making better-than-expected progress on its more advanced process nodes. This is critical to gain ground on TSMC,” says O’Donnell.
Newman, too, found Intel’s recent media blitz reassuring. “The accelerated roadmap indicates that the company is seeing execution progress, and given the stakes, I believe Intel’s certainty of its success is growing. It was one of the best moments for Intel in some time, and if it delivers, the market should finally come back around to the chip leader,” he said.
Some financial analysts have said 2023 will be the turnaround year for Intel, but Bajarin thinks it will start in 2024, with the real turnaround coming in 2025. “Unfortunately, a lot of the [economic challenges] are not in their control. It’s going to hurt AMD as well. There’s just nothing they can do. But they’ve got to manage through it,” he said.
Copyright © 2023 IDG Communications, Inc.