Microsoft and Nvidia are now the world’s most valuable companies. How Intel—once the top chipmaker—blew its shot to be one of them

[hist78]

"In the annals of business success stories few saw coming, Nvidia’s incredible rise to surpass Microsoft and become the world’s most valuable company is one for the books. This past week Nvidia, cofounded by Jensen Huang, edged out Microsoft as its market cap soared to $3.42 trillion. Yet it could have been otherwise. As business leaders marvel at Nvidia’s rise, it’s worth remembering that another company—Intel—came tantalizingly close to being where Nvidia is now.

Instead, Intel is waging a heroic effort to regain its place as the world’s preeminent maker of leading edge chips. The story of how it lost that position is part of a recent Fortune case study."

Microsoft and Nvidia are now the world’s most valuable companies. How Intel—once the top chipmaker—blew its shot to be one of them

As business leaders marvel at Nvidia’s rise, it’s worth remembering that another company—Intel—came tantalizingly close to being where Nvidia is now.

finance.yahoo.com

 

[Xebec]

Otellini totally missed the ball with Apple. The only thing I’ll say in Intel’s defense at the time was Windows Mobile was still viable at that time, and Intel was putting XScale in some of those devices around the time the decision was made (i.e. HP iPaq). Intel should have done something for Apple (in retrospect) but Intel’s (buying chips) partner Microsoft also failed hard around the same time.

2006 was a really mixed year for Intel: Bad decision on Apple, sold XScale at the end of the year, great success with Core 2, and won the Mac business for x86. Even AMD got into a mixed time period with buying ATi in 2006 and then selling the mobility Radeon IP to Qualcomm for a paltry $65M just 3 years later, right before Mobility really took off. 2006 is also likely when initial Bulldozer development started.

Intel also blew a strong and early lead in cloud that they had built by outcompeting other architectures in the late 1990s. There are actually many other ways Intel could have been a lot stronger today.

 

[meo9725]

Otellini totally missed the ball with Apple. The only thing I’ll say in Intel’s defense at the time was Windows Mobile was still viable at that time, and Intel was putting XScale in some of those devices around the time the decision was made (i.e. HP iPaq). Intel should have done something for Apple (in retrospect) but Intel’s (buying chips) partner Microsoft also failed hard around the same time.

2006 was a really mixed year for Intel: Bad decision on Apple, sold XScale at the end of the year, great success with Core 2, and won the Mac business for x86. Even AMD got into a mixed time period with buying ATi in 2006 and then selling the mobility Radeon IP to Qualcomm for a paltry $65M just 3 years later, right before Mobility really took off. 2006 is also likely when initial Bulldozer development started.

Intel also blew a strong and early lead in cloud that they had built by outcompeting other architectures in the late 1990s. There are actually many other ways Intel could have been a lot stronger today.

You cannot compare Otellini to Huang, or for that matter Jobs, or Bezos. After Moore and then Grove left, Intel was destined to falter. Intel ran out of great ideas and is unable to attract cream of the crop talent. Without these two ingredients it ain't gonna go anywhere.

 

[siliconbruh999]

You cannot compare Otellini to Huang, or for that matter Jobs, or Bezos. After Moore and then Grove left, Intel was destined to falter. Intel ran out of great ideas and is unable to attract cream of the crop talent. Without these two ingredients it ain't gonna go anywhere.

I agree difficult to compare grove and moore with now CEOs of Intel had gelsinger not left in 2009 wonder what would have happened

 

[BruceA]

You cannot compare Otellini to Huang, or for that matter Jobs, or Bezos. After Moore and then Grove left, Intel was destined to falter. Intel ran out of great ideas and is unable to attract cream of the crop talent. Without these two ingredients it ain't gonna go anywhere.

Where does one place blame for when leadership misses opportunity repeatedly?

Nvidia, AMD, Microsoft, and Amazon one could argue successfully pivoted. Great manager, leaders, board, culture?

Grove and Moore are legends but they failed miserably in seeding a board and succession plan. Intel really could and should be Nvidia and TSMC and AMD all together. If you add up their headcount it pretty much has that many people but not the results or vision.

 

[meo9725]

Where does one place blame for when leadership misses opportunity repeatedly?

Nvidia, AMD, Microsoft, and Amazon one could argue successfully pivoted. Great manager, leaders, board, culture?

Grove and Moore are legends but they failed miserably in seeding a board and succession plan. Intel really could and should be Nvidia and TSMC and AMD all together. If you add up their headcount it pretty much has that many people but not the results or vision.

You're right about the lack of strong succession plan. This had costed Intel dearly.

 

[osnium]

Where does one place blame for when leadership misses opportunity repeatedly?

Nvidia, AMD, Microsoft, and Amazon one could argue successfully pivoted. Great manager, leaders, board, culture?

Grove and Moore are legends but they failed miserably in seeding a board and succession plan. Intel really could and should be Nvidia and TSMC and AMD all together. If you add up their headcount it pretty much has that many people but not the results or vision.

I would blame Craig Barrett as opposed to Grove and Moore for a bad succession plan.

 

[BruceA]

CRB, PSO, BK what company could survive that ?

A board asleep at the wheel didn’t help with Andy really being terrible as well

 

[osnium]

CRB, PSO, BK what company could survive that ?

A board asleep at the wheel didn’t help with Andy really being terrible as well

Was Craig Barrett a bad CEO as well? I know PSO and BK were incompetent.

 

[nghanayem]

Was Craig Barrett a bad CEO as well? I know PSO and BK were incompetent.

CB was excellent. He rebuilt intel's manufacturing arm into a globally competitive execution machine before becoming CEO. Unlike some of his successors, he was willing to spend money on capacity expansions as well as on semiconductor R&D. He also made the excellent call to go hard and fast on 300mm; gaining a massive cost advantage over the rest of the industry even if you ignored the titanic process lead. The only black mark is that, if memory serves, the anticompetitive nonsense that the sales department participated in happened under his watch. The expensive side distraction business/failures also didn't start until PO/BK eras.

 

[osnium]

CB was excellent. He rebuilt intel's manufacturing arm into a globally competitive execution machine before becoming CEO. Unlike some of his successors, he was willing to spend money on capacity expansions as well as on semiconductor R&D. He also made the excellent call to go hard and fast on 300mm; gaining a massive cost advantage over the rest of the industry even if you ignored the titanic process lead. The only black mark is that, if memory serves, the anticompetitive nonsense that the sales department participated in happened under his watch. The expensive side distraction business/failures also didn't start until PO/BK eras.

Yes, I agree with your assessment. The other big knock on CB would be the succession issue.

 

[hist78]

CB was excellent. He rebuilt intel's manufacturing arm into a globally competitive execution machine before becoming CEO. Unlike some of his successors, he was willing to spend money on capacity expansions as well as on semiconductor R&D. He also made the excellent call to go hard and fast on 300mm; gaining a massive cost advantage over the rest of the industry even if you ignored the titanic process lead. The only black mark is that, if memory serves, the anticompetitive nonsense that the sales department participated in happened under his watch. The expensive side distraction business/failures also didn't start until PO/BK eras.

In retrospect Intel CEO Paul Otellini's decision to skip the opportunity to manufacture chips for the iPhones is relatively a smaller mistake in comparison with Craig Barrett's.

Craig Barrett's decision (or at least he was part of the Intel decision making team) around 1986 ~ 1987 to decline Morris Chang's invitation to be one of the TSMC's founding investors is a fateful mistake. Instead Phillips invested 28% of TSMC shares and helped TSMC to get off the ground.

At last Friday's (June 21, 2024) stock market closing, 28% of TSMC market cap would be US$252.62 billion. Intel's market cap on last Friday was $132.35 billion, a much humble one. Additionally, now officially Intel is entering the foundry business that Craig Berrett passed over.

 

[BruceA]

In retrospect Intel CEO Paul Otellini's decision to skip the opportunity to manufacture chips for the iPhones is relatively a smaller mistake in comparison with Craig Barrett's.

Craig Barrett's decision (or at least he was part of the Intel decision making team) around 1986 ~ 1987 to decline Morris Chang's invitation to be one of the TSMC's founding investors is a fateful mistake. Instead Phillips invested 28% of TSMC shares and helped TSMC to get off the ground.

At last Friday's (June 21, 2024) stock market closing, 28% of TSMC market cap would be US$252.62 billion. Intel's market cap on last Friday was $132.35 billion, a much humble one. Additionally, now officially Intel is entering the foundry business that Craig Berrett passed over.

Disagree here, Apple was the catalyst and primary driver to bring TSMC to advanced node, cadence as well as scale. Without them TSMC would have never caught Intel nor had the business foundation to develop and fund their technology. Apple volumes and cadences forced TSMC to put in place a cadence and the volumes guaranteed them ROI. You can read another’s prospective of that here.

The Apple-TSMC Partnership

The TSMC-Apple partnership has been one of the defining relationships of the mobile computing era.

semiconductor.substack.com

When PSO declined he gave up a demanding customer who would have driven development cadence and volumes that would have either pushed or distracted the company from its core bussiness. It was their first of two pivots missed by Intel ( mobile and GPU). Imagine if he did accept and Intel had delivered! They’d have had fabs at scale making x86 and SOC volume of half billion chips and Fab14, Fab15 and Fab18 wouldn’t exist and one could argue nor would AMD and Nvidia maynot be where it is at. They not have to do Foundry at all as they d have the volume of just two customers to drive technology and likely that customer would have forced LTD’s transparency on their 10nm.

The world would / could be worse for it and Intel far better. It is good Paul made that mistake and it was far greater a blunder from Intel’s prospective than any Craig made!

 

[hist78]

Disagree here, Apple was the catalyst and primary driver to bring TSMC to advanced node, cadence as well as scale. Without them TSMC would have never caught Intel nor had the business foundation to develop and fund their technology. Apple volumes and cadences forced TSMC to put in place a cadence and the volumes guaranteed them ROI. You can read another’s prospective of that here.

The Apple-TSMC Partnership

The TSMC-Apple partnership has been one of the defining relationships of the mobile computing era.

semiconductor.substack.com

When PSO declined he gave up a demanding customer who would have driven development cadence and volumes that would have either pushed or distracted the company from its core bussiness. It was their first of two pivots missed by Intel ( mobile and GPU). Imagine if he did accept and Intel had delivered! They’d have had fabs at scale making x86 and SOC volume of half billion chips and Fab14, Fab15 and Fab18 wouldn’t exist and one could argue nor would AMD and Nvidia maynot be where it is at. They not have to do Foundry at all as they d have the volume of just two customers to drive technology and likely that customer would have forced LTD’s transparency on their 10nm.

The world would / could be worse for it and Intel far better. It is good Paul made that mistake and it was far greater a blunder from Intel’s prospective than any Craig made!

I was talking about both Craig Barrett and Paul Otellini missed the foundry business opportunities. One opportunity was presented by Morris Chang in person to Craig and the other was from Steve Job to Paul Otellini. But neither Craig nor Paul is a bad CEO or the worst CEO. For a long time, TSMC was a not even a Intel's competitor at all. Why Intel kept missing opportunities like these two just mentioned?

In terms of how TSMC came to its leading position today, I believe it deserves several books, research projects, or a Semiwiki thread to explore. Apple and TSMC's partnership is one of several important factors. But it's not the only one.

 

[TICin852]

I was talking about both Craig Barrett and Paul Otellini missed the foundry business opportunities. One opportunity was presented by Morris Chang in person to Craig and the other was from Steve Job to Paul Otellini. But neither Craig nor Paul is a bad CEO or the worst CEO. For a long time, TSMC was a not even a Intel's competitor at all. Why Intel kept missing opportunities like these two just mentioned?

In terms of how TSMC came to its leading position today, I believe it deserves several books, research projects, or a Semiwiki thread to explore. Apple and TSMC's partnership is one of several important factors. But it's not the only one.

Why do you think Craig Barrett's opinion would have heavily impacted /influenced the decision? In 1986-7, when Morris Chang was looking around for funding, Noyce, Moore and Grove were still in charge.

 

[MKWVentures]

TSMC, Nvidia, Microsoft are just better companies than Intel. Intel was better than AMD, Motorola, Zilog is the 80s and 90s.
Intel has never done well at software, graphics, mobile or cost effective manufacturing. Even in the 90s. They are great at CPU design, massive ramps, and moving compute ecosystem forward at a methodical pace.

 

[hist78]

Why do you think Craig Barrett's opinion would have heavily impacted /influenced the decision? In 1986-7, when Morris Chang was looking around for funding, Noyce, Moore and Grove were still in charge.

Around 1986/1987 Craig Barret was the lead person at Intel to review Morris Chang's proposal to start the TSMC/dedicated foundry business. At the end of the second meeting Craig Barrett told Morris that Intel is not interested to join.

You can watch this interesting video below or just the part related to the Intel starting at 39:30.

 

[hist78]

TSMC, Nvidia, Microsoft are just better companies than Intel. Intel was better than AMD, Motorola, Zilog is the 80s and 90s.
Intel has never done well at software, graphics, mobile or cost effective manufacturing. Even in the 90s. They are great at CPU design, massive ramps, and moving compute ecosystem forward at a methodical pace.

I believe Intel's business model created a chain reaction that leads to today's Intel and its problems. Instead of changing its business model drastically to face the changing world, Intel wasted too much time to prove its business model is still valid and effective. At the same time the industries changed and moved on without Intel.

 

[MKWVentures]

I believe Intel's business model created a chain reaction that leads to today's Intel and its problems. Instead of changing its business model drastically to face the changing world, Intel wasted too much time to prove its business model is still valid and effective. At the same time the industries changed and moved on without Intel.

Perhaps. I tend to think companies are good at one thing or another and changing is not really effective. Its like saying Versace should be more like Ross to grow revenue. They could do a retraining and see what happens but why? Intel has many skills and tons of talent.

Also, Intel is valuable for what it is good at. Large scale product ramp and compute architecture. Once they stop trying to be good at what they have never been good at, they will be very successful again (I predict 2028 turnaround)

Can you imagine if Versace lost millions on their discount brands, and then said "this will be our new focus for the next 10 years, we will move away from $10,000 Designer outfits"... most of us would say "if I want Ross, I will go to Ross (or TJ Max)"

FYI the converse is true as well. Ross cannot be Versace.

Analogies like Race cars vs limos, 787 vs lear jet, Chik Fila vs Morton's, Opera singer vs Rock Singer, etc apply as well.

Just an opinion

 

[osnium]

Perhaps. I tend to think companies are good at one thing or another and changing is not really effective. Its like saying Versace should be more like Ross to grow revenue. They could do a retraining and see what happens but why? Intel has many skills and tons of talent.

Also, Intel is valuable for what it is good at. Large scale product ramp and compute architecture. Once they stop trying to be good at what they have never been good at, they will be very successful again (I predict 2028 turnaround)

Can you imagine if Versace lost millions on their discount brands, and then said "this will be our new focus for the next 10 years, we will move away from $10,000 Designer outfits"... most of us would say "if I want Ross, I will go to Ross (or TJ Max)"

FYI the converse is true as well. Ross cannot be Versace.

Analogies like Race cars vs limos, 787 vs lear jet, Chik Fila vs Morton's, Opera singer vs Rock Singer, etc apply as well.

Just an opinion

What would Intel look like in 2028 (the turnaround date)? Would it be a fabless company? Will they drop the foundry effort?