Pat Gelsinger's inevitable end

So Pat Gelsinger has “retired”.  I always felt that they would have to take Pat out of this job in a coffin, so this is not a voluntary retirement.  Pat faced a nearly impossible task; this end was probably inevitable for him.

There are two relevant newsletter reads that came out yesterday before the Gelsinger news.   Ben Thompson at Stratechery wrote about the AI transition happening in the industry.  Ben introduces a framework for thinking about applications and how each major hardware transition — from mainframe to PC, from PC to Phone, from Phones to maybe wearables — was bridged by an application model.  The read is a little dense, and even if you don’t buy it all, it stands out that Intel is nowhere in the discussion.  Intel allowed itself to get stuck on a hardware platform with no bridges to the next generation of hardware.  

Contrast this with Microsoft, which also faced the challenge of navigating across hardware transitions and application model transitions.  And Microsoft struggled with these transitions but has come roaring back.  There are a lot of moving parts in the Microsoft story, but three elements stand out to me:

• Microsoft was willing to walk away from Windows as the center of its strategy as it lost relevance.
• Microsoft spread Office everywhere, ensuring that Office remained front and center with users.
• Microsoft quickly (and at some risk) jumped into AI leadership through the OpenAI deal — Microsoft had struggled with the transition to mobile and would not be left behind again.

Contrast Intel also with NVidia — NVidia has invested heavily in its software stack and development tools and in cloud partnerships to ensure that they are deeply involved with the ongoing development and deployment of applications.   Intel spent much money on software but never really figured out how to make any part of the Intel software stack necessary for customers.

Bryne Hobart at The Diff has an interesting take on wealth creation around platform transitions.  He makes the strong argument that wealth is captured not so much by the creators of the platform but by the complementary businesses that grow alongside that platform:

What each of these large-scale booms has in common is that it starts with the capital-intensive growth of some universal complement—faster transportation, cheaper energy, quicker computation, more information—and then figuring out which adjacent business has durable competitive advantages.

In the PC Wave, Microsoft captured much of the wealth creation as the complement to low-cost chips and computers and did so with much less capital intensity.  Unfortunately, Intel took on much of the capital-intensive work in that wave.   Microsoft ended up being a nimbler company with assets that were easy to retarget and redeploy as additional platform transitions occurred.  

When you consider both these frameworks, Intel faced a very tough challenge stretching back to the 90s — they had assets that were hard to redeploy, and they were tied to the hardware for one particular platform wave.  Couple that with management missteps over several decades, and Pat Gelsinger faced an almost impossible task.  Which he probably knew, and I give him a lot of credit for trying.