Hi from San Francisco. Intel may need to adjust its definition of leadership if it expects to thrive in today’s chip industry. But first... Three things you need to know today: • Adobe underwhelmed investors looking for a moneymaking AI strategy • Microsoft and cyber firms are looking to prevent another CrowdStrike meltdown • Silicon Valley’s Y Combinator will double the number of annual sessions Intel Corp. Chief Executive Officer Pat Gelsinger has a strong sense of the company’s historic role in the computer industry and a keen understanding of the rewards of leadership. He confronted the consequences of losing that lead last month when announcing massive job cuts, a suspension of dividends and a reduced budget for his ambitious turnaround plan. “We were never built for efficiency,” he told the audience during a conference call following what analysts described as the company’s worst- ever report. “We were built for leadership.’’ This week, Gelsinger met with the board to figure out how to save money to keep his comeback plan on track. But, according to longtime observers of the company, Intel’s management needs to bring a more nuanced understanding of what leadership means and maybe refine their plans accordingly. When Gelsinger left Intel, a company he had joined as a teenager, to become an executive at EMC Corp. in 2009, the chipmaker was at the height of its powers. Its product releases set the agenda for the whole computer industry. Its processors were the essential heart of pretty much all servers and the vast majority of PCs. Its innovations defined the future direction of technology. Stratospheric market share — still abnormally high in the eyes of competitors such as Advanced Micro Devices Inc. — gave Intel massive profits and resources, which it used to maintain its central role. The company’s big spending on research and development and the engineering it did around its products made switching suppliers costly and risky. The payoff for Intel was always the widest profit margins in the industry. Rivals such as Nvidia Corp. and AMD lived in fear that Intel would take over their market niches. The prospect of joining the long list of those who had to give up even trying to compete against Intel was constant until just a few years ago. Now chip buyers have choices and, in the ultimate reversal, Intel needs those erstwhile enemies as it tries to become an outsourced manufacturer. Currently its so-called foundry operation almost completely relies on orders from its own designers. An organization used to defining customers’ products, charging high prices and spending freely to overwhelm competitors now needs to learn to sell itself as a partner. Intel has to convince a very skeptical group of companies who’re finally emerging from its shadow to trust it. Wooing rivals may be its toughest task. “They’ve all had their teeth kicked in by Intel for most of their career,’’ said Jon Bathgate a fund manager at NZS Capital in Denver. “Now they see a window to put the pedal to the metal.’’ Gelsinger’s answer to this longstanding question has been to argue that if he has unquestioned leadership (there’s that word again) in technology, his customers, and competitors even, will give him orders. Intel’s CEO has been bullish about the prospects of new AI PC chips. But Gelsinger conceded it will take longer to produce competitive products for servers and more importantly chips called accelerators for artificial intelligence — a category where Nvidia is raking in tens of billions of dollars of orders. But the pursuit of so many tough goals with a shrinking pool of resources is contributing to the delay in showing progress, according to Daniel Morgan, a fund manager at Synovus Trust Co. Intel’s revenue is down more than 30% from its 2021 peak, the debt is more than $50 billion, and cash flow has turned negative. Instead of pursuing multiple crushingly difficult goals aimed at restoring Intel to its omnipotence in the computer industry, the company should reduce its ambitions and focus on success in one area. A retreat from manufacturing to concentrate on design — a course taken by AMD more than a decade ago — is the best path, Morgan said. “You can’t continue to do multiple things and not do anything well,” said Morgan who’s been following and investing in Intel for decades. Trying for leadership in design and manufacturing, even in an industry where production technology adds better capabilities to chips, is too difficult, according to Morgan. “Used to be that Intel was the No. 1 — they totally dominated,’’ he said. “Now you have three or four other companies that are beating it.”—Ian King OpenAI is ready to take the next big step with artificial intelligence as it looks to stay ahead in the emerging field. The startup is releasing a model known internally as “Strawberry” that can perform some human-like reasoning tasks. The entire staff of video game studio Annapurna Interactive resigned this month following a dispute with its owner. Oracle says revenue will top $104 billion by fiscal year 2029, spurred by its cloud business. Tech leaders including Nvidia’s Jensen Huang met with Biden administration officials to discuss the massive equipment needs for artificial intelligence. French AI startup Poolside is nearing a $3 billion valuation to rival Microsoft’s GitHub. |