Should Microsoft Ride in to Save Intel?
As everyone is well aware, Intel is in trouble. Not in the imminent collapse sort of way, but in a way that's no less terrifying: Pat Gelsinger's turnaround plan clearly isn't working – at least not on the timetable that was promised – and there would seem to be no good solutions. Ian King sets the bloody stage in an in-depth look both backwards and forwards for Bloomberg:
It didn’t have to come to this. Intel’s strength in making chips for data centers should have left it well-positioned for the sudden rise in artificial intelligence. But it lagged in the race to produce the specific kind of equipment needed to train and operate AI models, and has almost entirely missed out on the recent boom. Intel is headed towards its third consecutive year of shrinking sales, estimated to make $52 billion in revenue in 2024, just 70% what it brought in back in 2021. Its shares have lost more than 60% of their value this year, turning them into the second-worst-performing stock on the S&P 500.
They've also, of course, had a GPU business for years. But just as with mobile, they seem to have been focused almost completely in the wrong direction. With graphics they were less focused on being the most powerful and instead being integrated and efficient. With mobile (and CPUs in general), they were too focused on being the most powerful and not being integrated and efficient enough.
None of that is Gelsinger's fault, of course – though he was at the company when such strategic decisions were made! – but his path to turn things around, which sounded good on paper, was always perhaps a bit unrealistic. The two reasons why are both obvious and the most important elements of any business turnaround: money and time. On time, as I concluded nearly four years ago, remarking on some of Gelsinger's turnaround comments upon being brought back:
Anyway, back to Gelsinger. The question now is if the prodigal son returning can right the ship. Is this akin to Steve Jobs returning to Apple and is this TSMC outsourcing akin to the Microsoft deal/investment? That’s probably a stretch and too tall an order, of course. And again, Intel is still in a better place than Apple was when Jobs returned.
Still, there’s a bit of an Anagnorisis feeling here. Like Blackberry back in the day. They’re not dead (yet), but they know they’re in real trouble. Which is a good first step, I suppose. If years late.
And this was before AI – and NVIDIA – started to take off in a major way. That just made the timing element more pertinent. Suddenly, Intel was under assault from all sides – CPUs, Servers, AI. As for money, back to King:
The bleak outlook made one thing clear: Intel’s existing businesses aren’t performing well enough to allow it to spend its way back to relevance. “While the overall strategy might have made sense at the outset, the current runway of the business doesn’t seem to give enough support to get it to the end anymore,’’ Bernstein Societe Generale Group analyst Stacy Rasgon wrote in a note last week. "Something clearly has to be done. But what?’’
The options the board is considering this week are intended to help Intel find a more solid financial footing, even if that means trimming its ambitions, according to people familiar with its deliberations, who asked not to be identified because the discussions are private. It’s not clear which ones are most likely, and all of the possibilities face real barriers. The board hasn’t received any offers from potential buyers for the company, in part or in whole, and has not scheduled any binding votes
The board doesn’t plan to discuss replacing Gelsinger, who most Intel insiders see as taking over a company after the real damage had already been done. “He did create the expectation that all the challenges would be resolved in a shorter time frame,” said Hendi Susanto, portfolio manager at Gabelli Funds, an Intel investor. Still, Susanto added, “it’s hard to come up with who could do a better job.”
Indeed. So I'm trying to think outside the box a bit here; what if there's a path forward by looking a bit backwards? Just as Microsoft was Apple's white knight back in the day, what if they step up to be Intel's here? Not buying the company – that would never be allowed, nor would Microsoft want that overheadache,1 I imagine – but as with Apple back in the day, what about a large cash infusion with some deeper partnership promises?
To be clear, this would need to be far larger than it was with Apple, but Microsoft also has a lot more cash and is far more cash-generative than it was back then. In our era of acquisition dead-ends, all of Big Tech is looking for interesting ways to put money to work – none so more than Microsoft. And there are compelling elements here, I think. In AI, everyone is now trying to break being beholden to NVIDIA. And Microsoft is one of those key, massive customers, thanks not only to their work, but the work of their key strategic partner, OpenAI.
Even Microsoft can't get enough access to NVIDIA's chips on its own and needs to barter with would-be rivals for more chips. They can't be happy about this – no large company can. And they have to be working to hedge this problem – all large companies must. Intel, while basically nowhere in AI, can still help with some blueprints Microsoft can perhaps provide. This wouldn't yield results anytime soon, but if the best time to do something was yesterday, the second best time is today. And the two sides already have some sort of partnership for Microsoft-designed chips. So this may well be that – but I'm talking scale here, not little tests.
In CPUs, Microsoft has tried to tie itself closer to Qualcomm, but it's not yet clear that this will be good enough to combat Apple's deeper in-house chip integration. Again, they wouldn't be bringing Intel inside here, as it were.2 But a deep partnership at arms-length – something, which again, Microsoft knows a thing or two about – could yield results, I have to imagine. Why not take a big swing to hearken back to the halcyon days of "Wintel"?
None of this would solve the larger issues that got Intel to its current state, but perhaps the deep, unfortunate job cuts will sort that part out. Waking Intel back up, in effect:
Intel’s grip began to loosen about a decade ago. This happened for a number of reasons, including an internal culture marked by increasing complacence and risk aversion; the failure of two consecutive CEOs to anticipate technological trends and keep pace with rivals; and a general shift of the semiconductor industry from the US to Asia. By the time Gelsinger took over, he acknowledged that it would take years of capital investment to regain the ability to make cutting-edge semiconductors. Only then, he said, would Intel be competitive enough to go back to focusing on profitability.
It's a cautionary tale that should force all companies – no matter how dominant in the moment – to look in the mirror. Back to the money:
Yet Gelsinger insists the time to reap the rewards is finally arriving. He says Intel’s foundry business has “engagements” with a number of customers for $15 billion, spread out over multiple years. (TSMC’s foundry business, by comparison, is estimated to bring in $88 billion in revenue this year.) Until Intel scores high-volume business from an industry powerhouse like Apple Inc. or Nvidia to build their most advanced chips, suspicions will remain that its foundry business mostly exists to make chips for Intel’s own semiconductor-design operation.
With enough capital, perhaps small wins can lead to bigger wins can lead to bigger wins, re-establishing a level of confidence that Intel has clearly been lacking. And while my scenario above would undoubtedly still take time to yield "high-volume" business for Microsoft, their cash and commitment could buy that time.
By Gelsinger’s own admission, Intel hasn’t developed a compelling way to elbow its way into AI-specific chips, the most important part of the semiconductor industry today. While it has a line of processors that compete with Nvidia’s core products, the CEO acknowledges that Intel isn’t going to be most customers’ first choice. The competition for chips that can train AI models, he says, is a four-horse race between Nvidia, AMD, companies designing their own in-house chips, and Intel. “Intel is No. 4,” he says. “That’s hard.”
Yes, it is. But one way around this may be to effectively become the "in-house" chips for one of the major players. Like Microsoft, for example.
And, by the way, while working on all this and keeping the peddle to the metal on CPUs (which does seem to be making progress), they need to also be coming up with the proverbial "what's next", so they don't again miss mobile and AI.
One more thing: the other big loser if Intel falls is the US government. Which has not only promised billions to the company, but has done so to try to ensure that the country has a credible solution and option in chips. That same government perhaps just helped Microsoft land a deal with G42 to help with tangential national interests. So they might actually look favorably upon Microsoft as the white knight here riding in to save Intel...3
M.G. Siegler
M.G. Siegler
M.G. Siegler
M.G. Siegler
M.G. Siegler
1 Did I just coin this term? Can I trademark it?
2 See what I did there?
3 Then again, I'm often trying to get Microsoft to ride in and save companies... But they should have with Twitter too! Imagine that data set for AI training purposes now, if nothing else!