Intel Inside, Gelsinger Outside
The first computer my family had when I was growing up was an IBM PC with an Intel 386 chip inside. Every single one of those chips, as it turns out, had the initials "P G" etched on them. That was a 24-year-old Intel engineer named Pat Gelsinger making his mark, quite literally.
While he obviously wasn't a founder of Intel, it's hard to imagine someone who devoted more of their life to the company, outside of perhaps Andy Grove (also not technically a co-founder, by the way, though there on the day of incorporation). And so when it was announced that Gelsinger was returning to lead the company he had joined as a teenager, there was obviously a lot of excitement. And some natural comparisons to Steve Jobs returning to Apple.
While the actual parallels there are blurry at best, the high-level notion of someone who has the company running through their veins seemed apt. If anyone could correct Intel, it was Gelsinger, so the thinking went. Of course, it was ultimately wishful thinking. But mostly, it seems, because while Intel may not have been on the verge of financial collapse as Apple was back in the day, it's arguably in a trickier position in a much more competitive and dynamic market, mixed with geopolitical realities that Apple did not have to deal with.
Put another way: Pat Gelsinger failed to save Intel because no one could save Intel. Certainly not in the three years he was given to do so.
Sean Hollister has a good overview of just what happened to put Intel in their current position. And most of it, of course, is not Gelsinger's fault. He was there for some of the decisions, but the buck did not ultimately stop with him until these past few years and while he may have let those bucks flow a bit too freely on certain projects while restricting them too much on others, those are hindsight quibbles. He was doing what he could to course correct the Titanic navigating in the seas of Europa. Good luck.
Should he have fully spun out the foundry business? Sure, maybe. It worked for AMD, but it's also a very different time and world than when that happened 16 years ago. And again, the geopolitical climate is such right now that America wants more foundries, not fewer. Gelsinger made a bet on the Biden administration's CHIPS Act, and it looked like it was going to pay off handsomely at various points, quite literally. But delays happen, both within large, bloated companies but for sure within the large, bloated government. And before Gelsinger knew it, targets were being missed, money was being trimmed, and Biden was out.
That combination, one suspects, may have been the proverbial straw the broke the camel's back last week, leading the board to push Gelsinger out. The time horizon kept extending, contracts weren't lining up, and a new incoming administration meant the promised money was less secure. And Gelsinger as the corporate face of that CHIPS act was perhaps going to be a political problem for Intel – even if the Trump administration ultimately will want the same thing: American assembly of the most important chips. They seem to prefer sticks (tariffs) to carrots (CHIPS Act) to make that happen.
At the same time, Intel was under assault from all sides. Simply making the wrong bet on mobile (read: not opting for efficiency over power) all those years ago may have been fixable, just as Apple's disastrous decision to allow for clones of their computers was all those years ago. But the AI Boom coming out of left field (the GPU world) just compounded the problems. As did the comeback of the aforementioned AMD. Servers also came under assault. And the leap of ARM designs from mobile to desktop (and servers!), not just with former partner Apple, but with "Wintel" sweetheart Microsoft cheating on them with Qualcomm! TSMC capitalizing on Intel's decade-ago investment in ASML to capitalize on all of the above. Not just a perfect storm of problems. A perfect shitstorm.
Gelsinger's solution was essentially that "the best way out is always through". Speeding up that ship to try to catch TSMC. It's one of those plans that sounds good on paper, but in practice, it's not clear why it would work. Was TSMC just going to sit back and get lapped? The reason they took over is because the game on the field changed. Again, first with mobile, then with AI. Another game-change was the only actual way Intel was going to comeback. And while such change is always inevitable, there's no real way to force such a change. The best you can do is plan for a future you think you can see, even if in a fuzzy way, and wait for it to come. But it could take decades. And you might almost die along the way.
This is basically the story of Jensen Huang and NVIDIA. They ultimately were in right place/right time for AI, but only after a few near-death moments. The reality is that Intel was probably too big and established to take such risks. As such, Gelsinger's plan was the inevitable one. The only one, really.
Where Intel goes from here is beyond murky. They clearly need some sort of visionary CEO who can come in and put in place a plan for where the world is going next. But they also have to execute on that plan and stay the course through the turbulence. It doesn't seem like Intel's current board is going to allow for that. Ultimately, fingers must be pointed beyond the CEO as well...
Just over a decade ago, I recall wondering the same thing about a then-floundering Microsoft. Ultimately, that led to the ouster of Steve Ballmer. But without a revamp of the board, one has to wonder if Satya Nadella would have been able to turn around Microsoft as he has – in spectacular fashion. If I'm Lip-Bu Tan or any of the other would-be candidates to run Intel, I need to make sure the board is in place to allow the company to succeed.
Something Steve Jobs learned the hard way.
One more thing: to beat the dead horse and bring this all back around, I'm still not sure some sort of real outside-the-box move isn't in order here. Like say, a major Microsoft investment in Intel. When Steve Jobs famously welcomed Big Brother Bill Gates, it was just as much about the partnership to bring Office to the Mac as much as the money (and to signal to the world that the "war" between PC and Mac was over). Here, the old partnership can be rekindled to save Intel and at the same time, help Microsoft diversify away from NVIDIA (and eventually perhaps Qualcomm for the Copilot+ PC initiative). To go full Apple and design their own chips to be made by Intel. While playing up the political angle to the government, perhaps alleviating some of the new antitrust tension. Just an idea!