Despite Intel’s recent woes, CEO Pat Gelsinger never expected to join some 15,000 of his colleagues in getting turned away at the door. Mr. Gelsinger, a renowned engineer and successful businessman, devised a sweeping rescue plan when he took the helm of the troubled semiconductor maker in 2021. Given the company’s long legacy of failures, it was never a quick fix. Gelsinger may represent Inter’s current downturn, but the problems began long before his tenure and will likely continue into the future.
How did Intel get here?
Mr. Gelsinger was tasked with addressing nearly two decades of bad decisions, all of which were exacerbated. As one half of the Wintel Alliance, Intel produced chips that worked with Microsoft Windows and became a behemoth that engulfed the industry. The huge profits from this partnership meant that the organization was reluctant to work too hard on new ventures that could distract from the company’s golden goose, which is still going strong years later. This means that there was some resistance.
In 2005, then-CEO Paul Otellini turned down the chance to make the iPhone a system-on-a-chip. This would have been easy for Intel, since it already makes XScale ARM chips for mobile devices. Popular mobile phones such as the BlackBerry Pearl 8100 and Palm Treo 650 are equipped with Intel ARM chips. A year later, the company sells XScale to Marvell, believing it can shrink x86 chips to work in smartphones. The first Intel Atom handsets showed some promise, but Snapdragon at the time was manufactured by a much smaller rival, Qualcomm, and easily beat them out.
At the same time, Intel was working on its own discrete GPU platform, Larrabee, based on the x86 architecture. Despite years of marketing bravado and proposals to “crush” AMD/ATI and NVIDIA, Intel ended marketing in 2010 in favor of bundling integrated graphics with regular processor products. The decision transfers a large portion of the GPU market to NVIDIA, which is expected to become a leading brand in gaming, supercomputers, cryptocurrencies and AI, with quarterly sales of $35.1 billion as of Nov. 20. .
Could Intel have foreseen the rapid rise of AI? Probably not. However, Reuters reports that former Intel CEO Bob Swan turned down the opportunity to invest in OpenAI in 2017. The company was looking for hardware partners to reduce its dependence on NVIDIA, and was offering generous deals in the process. But Swan said he didn’t see a future in generative AI, and Intel’s data center division reportedly refused to sell the hardware at a discount.
Intel’s core strengths have been in engineering quality, product reliability, and staying close to the cutting edge. (There are parallels between Intel and Boeing, both of which have seen their reputations for quality eroded in real time.) Sadly, Intel failed to produce 10-nanometer chips. As a result, Intel’s main business was in trouble. Expected deadline in 2015. The company’s famous “tick-tock” strategy involves launching a new chip process in one year, launching an improved version in the next, and shutting it down.
These issues allowed Intel’s competitors to step in and take away the competitive edge with more modern chip architectures. AMD held just over 10% of the chip market for much of the 2010s, but has doubled its market share in recent years. The biggest beneficiary, of course, was TSMC, the Taiwanese chip factory that became the envy of the world. Even though Intel controls most of the x86 processor market, it’s TSMC that makes chips for Apple, Qualcomm, NVIDIA, AMD, and others. Intel, on the other hand, had an outdated chip manufacturing process and was unable to catch up with its rivals.
gelsinger theory
Mr. Gelsinger is, as you might expect, close to Intel’s “rivers,” joining the company at age 18 and rising to the position of chief technology officer by 2001. He left Intel in 2009 to become COO of EMC, where he served as CEO. I’ve been using VMWare for about 10 years. After taking the reins of Inter, he laid out detailed plans to achieve its glorious comeback.
Step one is to separate Intel’s design and manufacturing business into two separate entities. Gelsinger highlights U.S. subsidies through the Biden administration’s Chips and Science Act, promising to build two new chip factories leveraging the same EUV (extreme ultraviolet lithography) technology used by TSMC. did.
Mr. Gelsinger was also determined to reestablish discipline in Intel’s chip business and return to a “tick-tock” structure. Unfortunately, production delays have been piling up since 2015, so Gelsinger’s only goal was to return to parity. In the meantime, Intel plans to rely on TSMC to manufacture some of its latest chips, which, while costly, could help the company address concerns that it is falling further behind.
While no one doubted the scale of the challenge facing Gelsinger, there was plenty of room for optimism. Gelsinger humbly accepted that Intel could not continue on its current path and had to accept its new position. He suggested that Intel could laugh through the short-term pain for the company’s ultimate benefit. If Intel can build for the future, leverage rivals to stay competitive, and regain trust in its processes, it will emerge as a winner from there. All I needed to do was not make anything worse.
The situation got worse
In late October, Reuters reported that Gelsinger made an outrageous gaffe when talking about TSMC. The CEO is quoted as saying, “We don’t want to put all our eggs in the basket of our Taiwanese factory,” and “Taiwan is not a stable place.” This angered TSMC so much that it ended the discount Intel had been using for years.
Sadly, Gelsinger’s desire to restore discipline to the chip sector backfired, with the latest Core processors marred by voltage instability issues. Intel was forced to extend warranties on these chips, which came at an additional cost that it couldn’t really afford. In August, the company posted a loss of $1.6 billion and pledged to cut 15,000 employees in a restructuring effort. But three months later, it was forced to post its largest quarterly loss in history with a loss of $16.6 billion, much of which was tied to a revaluation of company assets and layoff payments. Ta. To make matters worse, Intel’s new manufacturing process, 18A, reportedly failed critical tests ahead of its 2025 debut.
Perhaps Intel’s lowest point this year was when its stock price fell enough to make it an acquisition target. There were rumors that Qualcomm was potentially eyeing an acquisition, while there were also rumors that ARM had approached Intel about buying Intel’s product division.
Where does this leave Intel?
The New York Times reported that Intel’s board became frustrated with Gelsinger because his rescue plan “did not show results quickly enough.” But Intel didn’t intend to hire Gelsinger in 2021 and suddenly bounce back in 2024. Building a large, complex chip factory is not easy. Nor can they mobilize thousands of engineers to solve tough chip yield problems. And clearly, the decline that began in 2015 could never be reversed overnight.
Intel’s board of directors is currently searching for a full-time successor to Gelsinger, but it’s hard to know if anyone else would do anything different. At the end of the day, companies still need to build factories and fix processes to own and control their futures. Unless, of course, the next CEO is told to stop the bleeding and keep the money flowing, Intel remains the biggest player in the world of x86 chips, even deeply scarred after several bad quarters. name and will continue to make money for years to come.
It’s easy to imagine Intel’s board sitting idle in favor of a few years of healthy profits at the expense of the company’s long-term future. The company could continue selling modified versions of its existing desktop chips and cede technological leadership to the likes of AMD and Qualcomm. There will probably be 10 or 20 years of large industrial customers who will be happy to use Intel processors in their hardware as long as they use Windows. Perhaps that’s appropriate, given how big and rigid Intel has become, admitting it can’t move fast enough to evolve.
Given Intel’s extensive role in global technology, such a scenario probably won’t be allowed to happen. Even if the next administration criticizes the CHIPS Act (Intel is still slated to be the biggest recipient of funding), the assets of having a domestic manufacturer of Intel’s size are enough to keep some sensible government from falling. I won’t allow it. However, simply replacing the CEO will not suddenly solve the company’s big, intractable problems. It wasn’t Pat Gelsinger who screwed up the Raptor Lake power supply design, nor did we give him the opportunity to build the iPhone’s CPU all those years ago. He can own TSMC, but while the CEO sets the direction, a company the size of Intel can’t micromanage every process. So whoever replaces him will likely face a similar slew of issues to grapple with, knowing the board’s patience will be even shorter this time around.
