Lip-Bu Tan: Intel’s New – And Maybe Last – CEO

In his letter to Intel employees, new chief executive officer Lip-Bu Tan, who starts his new job next Tuesday, tells them that he is “never deterred by challenges.” And that is a good thing because cleaning up the chip maker and chip designer is indeed a Herculean challenge.

Ironically, Tan will take the helm at Intel on the same day that Nvidia-co-founder and chief executive officer Jensen Huang will take the stage as an AI rockstar at the latest GPU Technical Conference – and as the head of the largest and most profitable semiconductor manufacturer in history. Nvidia is only the fourth company to break $100 billion in sales in the IT datacenter market, something that might have been Intel if it had not been so badly managed in both chip products and chip manufacturing processes in the 2010s.

That is all water under the bridge at this point.

We do not do a lot of first person narration here at The Next Platform, but it is warranted here.

I don’t know Tan, but I would have had the coronavirus not put the kibosh on our Next AI Platform event scheduled for March 2020, where Tan was set to be a keynote speaker. My wife, Nicole, does know Tan, having spoken at a conference hosted by Walden International, the venture capital firm that Tan founded back in 1987. She reported at the time – and reconfirmed during our morning coffee as the sun was rising today – that Tan is bright, meaning energetic as well as smart, and is interested in what other people think as well as being an interesting person with which to have a lively and intelligent conversation. There is no higher praise in our household.

I cannot see how I would not enjoy the company of a man who named his company after Walden; or, Life in the Woods, published in 1854 by Henry David Thoreau, taking five years to sell 2,000 copies (at $1 apiece). Walden remained in obscurity until the beatnik generation had its influence on academia in the 1950s and early 1960s and a renewed sense of environmentalism and individualism swept over America in the late 1960s and 1970s. And suddenly, we were all reading Civil Disobedience, but also being told to behave ourselves, and then graduating into Walden. Which most assuredly teaches us to be contrarian.

My very old hard-bound copy of Walden, printed in 1942 and traveled much in my SC backpacks, and a round stone from Walden Pond near the bean field behind where Thoreau’s cabin was.

Walden is the book I keep a copy of in the beater truck, the Jeep, and the Challenger – just in case I get trapped somewhere for any reason. It has so many pithy one-liners, acute observations, and tender sensibilities that it holds you from beginning to end. There is nothing else like Walden, which, as the Walden VC site quotes on the very first page for startups to “advance confidently in the direction of your dreams,” as Thoreau put it.

There may be no one else like Lip-Bu Tan, and it could turn out to be fortunate for Intel.

Tan was raised in Singapore and got his physics degree from Nanyang University, and then went to MIT to get a master’s degree in nuclear engineering. The Three Mile Island nuclear accident put the kibosh on that, and so Tan headed west to the University of San Francisco to get an MBA. Tan invested heavily in Asian startups, notably sound card maker Creative Technology and graphics chips maker S3 Graphics. Tan was an early and enthusiastic investor in Semiconductor Manufacturing International Corp, the indigenous foundry of China, when it was established in 2000.

In 2004, struggling chip design software maker Cadence Design Systems added Tan to the board, and he was named co-CEO in October 2008 and president and CEO in January 2009. That turnaround was a tough one, and this one will be tougher. To more fully quote Tan from the employee letter:

“One of the things you will learn about me is that I am never deterred by challenges. Throughout my career, they have motivated me to solve hard problems. As I prepare to come on board, I believe we have a truly unique opportunity to remake our company at one of the most pivotal moments in its history.

That’s not to say it will be easy. It won’t be. But I am joining because I believe with every fiber of my being that we have what it takes to win. Intel plays an essential role in the technology ecosystem, both in the US and around the world. And, together, I’m confident we can turn our business around.”

There is no hint in this letter about what Tan will do, of course. He will need to talk to a lot of people inside Intel to get a sense of what its experts really think, and then plot a course for the future. How different this is from the one that Pat Gelsinger, his predecessor and long-time employee of Intel, remains to be seen. Gelsinger was Intel’s first chief technology officer, the inventor of the tick-tock method of manufacturing: change process first, perfect it, then change microarchitecture next with a slightly improved process, but do not try to do both at the same time. Gelsinger arguably should have been its CEO even before the Great Recession hit in 2009, when he left for EMC and then VMware before coming back to Intel as CEO in 2021. Had Gelsinger been tapped as CEO in 2009, perhaps Intel would be the fourth $100 billion IT company already, well ahead of Nvidia. Perhaps not.

Gelsinger seemed unable or unwilling to make the deep cuts necessary at VMware to make it a more profitable company, and we think he was also hesitant to make any deep cuts in the Intel employee base to free up cash so it could be pumped back into the foundry business. The reason is that cuts cost money, too, and with 18A and 14A processes coming online in 2025 and 2026, closing the manufacturing gap would close the product gap, at least in server and desktop CPUs. For whatever reason, Intel’s board ran out of patience with Gelsinger, and he was ousted in December 2024. Intel’s datacenter products chief, Justin Hotard, has left to be the CEO at Nokia, but desktop products chief (now CEO of the entire Intel Products group) Michelle Johnston-Holthaus, is staying on to run the whole shebang. Dave Zinsner, who was interim co-CEO with Johnston-Holthaus, goes back to being chief financial officer.

As soon as it was spun out of Dell, VMware open to acquisition by Hock Tan, founder of the company at the heart of the Broadcom conglomerate, and a man who is not afraid to make cuts. As the new Intel CEO, Lip-Bu Tan may not have any choice but to cut deeper into Intel, and may have no choice but to do a deal with Broadcom for Intel Products or with Taiwan Semiconductor Manufacturing Co for Intel Foundry. Perhaps the idea is to line up funding from the US government, with president Donald Trump saying in the past that he wanted to set up a sovereign wealth fund akin to what the oil rich Arab nations have. Perhaps Trump wants to replace the CHIPS Act with something bigger. If so, we will find out soon enough, and how the national security interests that the United States play into this.

For now, the questions are these: Will Lip-Bu Tan make cuts, or cleave the Intel company into two and be done with it, as has been rumored? And having done that, will the pieces be sold off – Intel Products to Broadcom and Intel Foundry to TSMC? Or will Intel, through the magic of a US sovereign wealth fund, be in a position to buy TSMC’s operations in the United States and keep the Intel name should China invade Taiwan?

Who knows?

A breakup doesn’t seem to be part of Tan’s thinking, and that might be happening because Tan agreed to take the CEO job at Intel and is already steering the process away from that option. It sure looks like it from Tan’s letter to employees:

“We have a chance to do something special together. In many ways, we are the founders of “The New Intel.” We will learn from past mistakes, use setbacks to strengthen our resolve and choose action over distraction to reach our full potential.

Together, we will work hard to restore Intel’s position as a world-class products company, establish ourselves as a world-class foundry and delight our customers like never before. That’s what this moment demands of us as we remake Intel for the future.”

Even if Intel is righted, TSMC is still in trouble and is investing $165 billion in foundries in the United States to hedge against a Chinese invasion.

Given the geopolitical situation, with a trade war going on between the US and China and tariff barriers going up higher and higher, we think the world needs at least two advanced foundries. And they may end up being two halves of a TSMC, each conjoined with another foundry.

One half might be Intel Foundry operations in the United States, Israel, and Ireland, plus TSMC foundry operations in the United States, Germany, and Japan; some months or years hence we might call this agglomeration American Semiconductor Manufacturing Co, headquartered in Santa Clara with development fabs in Hillsboro and Phoenix. You might call it Intel should China invade Taiwan. You certainly could not call it Semiconductor Manufacturing International Corp. . . . <Smile>

The other half of TSMC might be conjoined with the actual Semiconductor Manufacturing International Corp, headquartered in Shanghai with fabs located there and in Beijing, Tianjin, and Shenzhen from SMIC and from around Taipei, the capital of Taiwan, which is the heart of TSMC. One might call this agglomeration Xinhua Semiconductor Manufacturing Co, headquartered in Beijing.

What do you think Donald Trump and Xi Jinping have talked about? We are pretty sure it is not golf, and we are also pretty sure that everything is a deal and that they both believe this. Eventually, there will be a deal.

That is all many moves ahead from naming a new CEO at Intel, of course. And even if none of that scenario above comes to pass, by some sort of political miracle, Intel still needs to be righted and Tan is the man for the job. We look forward to hearing about the plan.

If you have built castles in the air, your work need not be lost; that is where they should be. Now put the foundations under them.” – Henry David Thoreau

Sign up to our Newsletter

Featuring highlights, analysis, and stories from the week directly from us to your inbox with nothing in between.
Subscribe now

3 Comments

  1. Lip-Bu Tan seems like a good person to fix Intel. He needs to make sure Intel Foundry customers are treated fairly. He needs to find people who can make sure the 18A and 14A processes get good yields. He also needs to find new leadership for Intel’s datacenter group. The recent Granite Rapids processors have Intel On Demand (a terrible idea) and no Xeon Max with HBM. To support running large language models locally, Intel should add the option of High Bandwidth Flash (HBF) to their client, workstation and server processors.

    Having 8P + 16E cores in a client processor is a mistake. It is too difficult to optimize software for that configuration because the P and E cores have different size caches. Apple’s M3 Max has 12P + 4E while the M3 Ultra has 24P + 8E. I think Apple’s configurations are much more sensible. Intel’s desktop processors should support MR-DIMMs.

    There are currently no Intel client or workstation systems with Thunderbolt 5. Intel invented Thunderbolt 5 yet Apple is the first company to bring it to market. The Mac Studio has six Thunderbolt 5 ports and the MacBook Pro has three Thunderbolt 5 ports. Thunderbolt 5 should be built into the chipsets for Intel’s client and workstation processors or integrated on the processors.

  2. Me opinion: If Intel Lip B. T. CEO dont its firts company objetive, great “FOUNDRY”, superior foundry… their future is dark.

  3. Definitely! Build castles in the air, put foundations under them, and do not, under any circumstances, rain or shine, win or lose, yin or yang, tic or toc, steam-cooked square patties or flame broiler burgers, great taste or less filling, hawk or dove, ebony or ivory, ketchup or ranch, cash or credit, spelling bee or spellchecker, quality or quantity, do not, under any of these circumstances, and all others, do not, for any reason whatsoever, because you’ll regret it for a long long time to come, do not, I repeat, do not, ever, miss the moat! ( https://www.nextplatform.com/2023/05/04/missing-the-moat-with-ai/ ) 8^b

Leave a Reply

Your email address will not be published.


*


This site uses Akismet to reduce spam. Learn how your comment data is processed.