Like many Intel employees, I was full of hope when Pat Gelsinger re-joined in 2021. The prodigal son and technology savant returned to the company he loved, and would put a capstone on his career by restoring Intel to its former greatness. It’s hard to describe how much an effect Pat’s initial return had on the company which for years had become risk-averse and overly financialized. In the beginning nearly everyone at Intel believed in him; not only because they thought his strategy of building factories for external customers could succeed, but also because they simply wanted the company to do something ambitious again. No longer — Pat’s dream for Intel has ended. This week I discuss why Intel failed under his leadership and where the company goes from here.
But first, if you are a paid subscriber you received a notification on Monday morning at 7:59 AM that I was selling my Intel shares. At the time the stock was up 5% (in my opinion irrationally) on the news of Pat’s departure. Since that message was sent, the stock is down nearly 20% while the rest of the market has been on a tear.
Secondly, a confession. I loved Intel. It was my favorite job ever; I met my best friend and now wife while there; and I met the most amazingly intelligent life-long friends and colleagues at Intel. This personal history led to irrational thinking about the stock, and wanting or perhaps even willing for it to go up. I should have sold it all when I was laid off — regardless of my opinion on its valuation and future prospects. This article is not the view of a disgruntled employee (I held the stock until this week, and was laid off over a year ago), but rather a letter from one whose hopes and aspirations for what could have been were never fulfilled.
I apologize if anyone bought Intel stock or did not sell it after reading any of my writings, particularly Intel Earnings Pushes the Stock Down 12%. Emotions clouded my judgement, and this will not happen again.
“Don’t ‘fall in love’ with an investment - be situation dependent and opportunity driven.”
—Charlie Munger
A Job No One Wanted:
When Pat Gelsinger took the CEO job in 2021, Intel was being run by a caretaker CEO (Bob Swan) who took the reigns after the CEO before him (Brian Krzanich) nearly destroyed the business through his negligence, and was finally forced out because he had sex with one of his employees. The search for Pat took over 2 years — a testament to how few truly capable people wanted the job. The reason for this is not just because of the aforementioned chaos, but more so because Intel began to lag in its core competencies — TSMC took the lead in manufacturing and AMD was at parity in design. A turnaround would be a gargantuan task, one which had a significant chance of failing. Enter Pat’s ambitious dream.
Pat received a mandate for change when he joined because the board and employees knew the company was broken and thus needed to do something drastic to revive itself. Here was a leader who was perceived to truly care about the future of Intel because he had spent the bulk of his career at Intel, went off to be chief executive at another company (EMC/VMWare) when Intel leadership wouldn’t make him CEO, and still came back to try to solve our problems when things were worse than ever — in short, the anti-Krzanich that employees had longed for. The Intel board of directors voted unanimously to back Pat’s plan to double down on manufacturing when he re-joined in early 2021, even though the path ahead would be long, expensive, and difficult. Despite this, Pat was let go before being able to see his plans come to fruition. What led to this failure and early departure? Current online speculation includes:
The 18A process has undisclosed technical/yield problems. “I’ve bet the whole company on 18A” says Pat on multiple occasions.
18A is good technically, but Intel can’t get big customers to commit to it. Why take a risk on Intel (who has been terrible at execution) when TSMC has been nearly flawless? Pat cites geopolitical risk from Taiwan, but this hasn’t converted potential customers. Even if a customer commits, Intel will have to give them a huge discount to take a chance on Intel, and thus financials will be rough through at least 2030.
The board wants to sell off the FABs because they are costing more than anticipated and will not be profitable until 2030 (short-termism to juice the stock). Pat disagrees and has resisted, but the board is unanimous against him.
Pat lost sight of the product groups, which fell behind, because he was overly focused on foundry. Intel products, particularly in the data center, have not benefited from the AI boom. Unforgivable for a technology leader.
Board members (especially Lip Bu Tan) wanted Pat to be far more aggressive in cost-cutting, but Pat resisted.
The stock performance is so horrific (and nearly all competitors are ripping higher) that the CEO must go.
No one yet knows for sure which combination of these reasons led to Pat’s downfall, so what I want to discuss are the precursors to these even being possible under his leadership.
These factors fall into three categories—
Failure to change Intel’s culture
Failure to incentivize employees
Failure to build trust with stakeholders
Confronting New Realities — Intel Culture:
Pat joined at 18 years old and worked at Intel for the next 30 years of his life (1979-2009). He was mentored by and worked closely with Andy Grove, one of the heroes of Intel lore. For anyone who’s read the fantastic books by Andy Grove (the most famous being “Only the Paranoid Survive”), it gives the feeling that Intel culture at the time was one which by today’s parlance would be described as “extremely hardcore.”
Keep reading with a 7-day free trial
Subscribe to Dragon's Den Investing to keep reading this post and get 7 days of free access to the full post archives.