Due terzi compiuti

Pensieri e esperienze di un’anatra selvaggia di ottanta e un anni.

Avram Miller by the sea
EN | IT

587 post pubblicati dal 2006

Tag: Andy Grove

  • Intel Must be Saved

    I write often about Intel and in particular is spectacular rise and fall.  I have explained that Intel under the leadership of Andy Grove, recognized  the opportunity of the Personal  Computer and brilliantly exploited it.  That as act two for Intel which had originally focused on memory devices. The PC opportunity in which Intel played not initial role in its creation came along just in time to save the company for bankruptcy or a sale to the Japanese (Grove once confided in me that this was something he saw as a possibility).  But Intel missed act three, mobile computing which created the opportunity for TSMC (Taiwan Semiconductor Manufacturing) to become the leading manufacture of semiconductors and to over take Intel in terms of technology in starting in 2017 and asserting a clear leadership position by 2020.

    Intel was/is a vertically integrated semiconductor company meaning that it both designs and manufacture its chips.  Companies like Apple and Nvidia (and many more), design chips but then utilize TSCM to manufacture them.  Apple made the decision to move the manufacturing of the processor for the iPhone to TSCM in 2014.  Steve Jobs, the CEO of Apple, approached Paul Otellini  the then CEO of Intel in 2005 about providing the processor for the yet to be designed iPhone.  Otellini famously turned Jobs down because he did not see the potential of the smart phone and was not interested in having a low margin business while Intel factories were operating at full capacity. Apple turned to Samsung.  The iPhone was introduced in 2007. In 2014, Apple moved to TMSC. Apple is TMSC largest customer, accounting for 20-25% of sales, followed by Nvidia and AMD.   Companies like Amazon, Google and Microsoft design their own chips and then have them made by founders like TSMC.   They probably represent about 20% of the semi conductor demand.  The equipment used by TSMC and other foundries is made by a number of companies with the leader being the Dutch company, ASML which has about 25% of the world market.

    When I joined Intel in 1984, the company was vertically integrated. It developed its own manufacturing capabilities and designed it own chips.  In 2000 for instance, half of Intel’s engineers were involved in manufacturing technology. The manufacturing technology group was called TMG.  It was initially led by Craig Barrett who become Intel’s COO in 1990 and succeeding Andy Grove as CEO in 1998.  He became board chair in 2005.

    It was Barretts appointment to CEO the cemented my decision to leave Intel which I did in April of 1999.  Intel was very insular during my time there (1984-1999).  Of the 33 Corp officers listed in the 1998 Annual Report, I was the last to have joined, 14 years earlier.   Andy, and I often spared about strategy.  I regretted my inability to have much impact on his thinking.  All of this is covered in my book, The Flight of a Wild Duck.  

    During my last years at Intel, I often had debates with Andy about the future of the PC industry.  At that time, I was leading Intel’s efforts to create residential broadband as well leading the consumer investments of Intel Capital.  Andy, realized that Intel needed to expand the companies business into other areas because the PC but was incapable of making that happen.  He thought that Barrett would be the one to do that but alas that was not the case. I use to say that Andy wanted the same old ideas from the same old people and the Barrett wanted new ideas from the same old people.  The was very clear when he took Gerry Parker the then leader of TMG and made him head of new business development even though Gerry never ran a business in his life.  The result of billions of dollars wasted in new business development and a reduction in leadership of TMG which proved deadly.  By the way, I offer Gerry my help in his new role but he declined.

    It is amazing to me that Grove who wrote a brilliant book, “Only the Paranoid Survey” in which he explained how the computer industry went from vertical integration with companies like IBM and Digital Equipment where companies had proprietary hardware and software as well as there own peripherals and service organization, to the PC world where different companies produced the hardware, peripherals, software and services.  What happened to Intel was describe  by Clayton Christensen in 1997 in his very impactful book, The Innovators Dilemma. Andy was well aware of the concept as we had Christensen teach a class to the executives and advise the company. This book explains a great deal about what happened to Intel and so many other companies.  It is hard for successful companies to reinvent themselves.  They cling to their success protecting business that generate high profits.  Talent manager seek to run the successful businesses rather than take risk to run new ventures.  And those that want the new ventures often leave to start their own successful business.  

    The story of Intel would just be another example if it was not for the strategic importance of having USA domestic leader in chip development.

    During my time at Intel. the PC boom drove Intel’s volume.  While the X86 architecture was not ideal, intel’s ability to push performance by outstanding semiconductor process development, allow the company to monopolize the PC industry.  

    What went wrong

    I blame Intel’s board, When I first joined Intel in 1984, the company had an amazing board. but it was a board that change under Grove’s chairmanship.  It became focused on governance and financial issues. Few on the board understood the semiconductor business and the importance of investing in the process technology.  After Andy’s chairmanship (1997-2005), Barrett become the chairman (2005-2009).  You would have thought he knew better but he did not actually have a good relationship with the other board members. He was succeeded by Jane Shaw (2009-2012) a health care executive, then Andy Bryant (2012-2020) the former CFO of Intel.  Bryant is in my opinion bares the most responsibly for the decline of Intel.   After Bryant (came Omar Ishrak (2020-2023) of Medtronic, another health care executive and finally Frank Yeary (2023-Present)  and investment banker.  

    Stock Buybacks

    One of the unforgivable acts of the board was supporting stock buybacks.  Intel used $108 billion dollars to by back its own shares as an average price of $45 dollars.   This is more than the company is currently worth.  Part of the downward  pressure on the stock results from concerns that the company does not have enough capital to pull of it strategy having a major boundary business.  

    After Grove left as CEO, their was a succession of CEOs, each worse than the previous.  One of the major responsibilities if not the main responsibility of a board is to appoint the right CEO.  Intel’s board failed spectacularly in this regard.

    . 1. Craig Barrett (1998–2005)

    2. Paul Otellini (2005–2013)

    3. Brian Krzanich (2013–2018)

    4. Robert Swan (2019–2021)

    Then there was Pat Gelsinger (2001-2004)

    Pat first joined Intel at the age of 18 in 1979.  I meet him soon after I joined Intel in 1984.  I remember being impressed by his energy.  He became one of Grove’s “fair hair boys”.  Since I was not really that involve with the microprocessor development, I did not have much to do with Pat until Pat was appointed by Andy to lead the failed video conferencing product,  Pro Share project. It was a poorly conceived but it was one of Andy’s failed to create to create new applications for PC market. Andy tapped Pat to run that project although Pat had no qualifications.  At that time, I was driving Intel initiative to create residential broadband.  I knew a great deal about communications and new that ISDN and working with the phone companies would led to a disaster but neither Pat nor Andy would listen.  Pat left Intel in 2009 to join EMC Corp as President and COO.  I believe that Pat was push out by Otellini, the CEO.  I was no longer at Intel and have no direct knowledge other than hearsay.  

    In many ways, Pat was a great choice to try to retrieve Intel.  But right away, I began having concerns.  One of Pat’s first pronouncements about his joining was to say “Intel is back.”  He kind of had a messianic tone to his speeches.  He like to liken himself and his return to Steve Jobs.  But Gelsinger was no Steve Job.

    Never the less, I think he did have a strategy for the company.  Pat recognized that need for there to be a foundry based in the USA to compete with TSMC especially given the geo political issues between Taiwan and China.  But building a boundary business is very capital intensive.  Intel would have to match or exceed the abilities to TSMC.  That would take years and billions of dollars. There problem is that Intel is a public company and investor in such companies are very short term oriented. 

    On Dec 2nd, Gelsinger abruptly resigned.  It is not really clear what happened.  Clearly there was a major disagreement about direction at the board meeting.  Over time, we are likely to learn more.  

    What will happen now

    I think there are a few scenarios.  The best would be for Intel to be taken private, perhaps by a consortium led by Michael Dell (Dell Computer) which would bring Pat back to lead this effort. 

    Otherwise, I think Intel will be sliced and diced and cease to exist as an independent company.   I hope that does not happen.  The USA and the world needs a vibrant Intel.  Intel is the only USA company that can build a capability similar to TSMC.  

  • Was the Personal Computer Industry Inevitable?

    While writing my book, The Flight of a Wild Duck, I couldn’t help but speculate on alternative realities. So much of one’s life is driven by random events. So it was easy for me to imagine how things could have unfolded differently. One of the possibilities that intrigued me was derived from the significant shift in the computer industry resulting from the introduction of the IBM Personal Computer in 1981, mainly because this had a profound impact on my career.
    Before its introduction, the computer systems industry comprised Mainframe Computers, with IBM having the majority of the market and Mini Computers with Digital Equipment Corporation. In addition, personal computers were a tiny market serving hobbyists and computer enthusiasts.


    In 1980, Apple had started to develop a business market driven by the introduction of VisiCalc, the first spreadsheet. IBM noticed and decided that they wanted to participate and even dominated the nascent personal computer business market. However, in their haste to get to market quickly, in less than 12 months, they failed to protect the architecture of their personal computer, which allowed companies like the newly formed Compaq Computer to develop fully compatible computers. IBM saw the personal computer as a new opportunity but failed to understand it would evolve and eventually take away market share from Mainframes and Mini Computers.IBM was not alone in this. No major technology company foresaw the titanic shift that would turn the computer industry on its side.


    IBM, the dominant player in the computer industry, began to suffer significant losses and had to exit the personal computer market and eventually restructure as a service company. In addition, minicomputer companies like Digital Equipment started to suffer substantial losses. Ultimately, Digital was acquired by Compaq, one of the leaders in the personal computer market. The driver of the personal computer industry was standards, mainly when it came to application software. The result was that PC manufacturers had difficulty differentiating themselves from one and another. Lead by Compaq Computer Corporation, manufacturers of IBM-compatible computers, known as clones, successfully competed against the very personal computer market that IBM had created. Eventually, Asian manufacturers entered the market with lower-cost versions, putting great pressure on IBM and American clone manufacturers.


    Microsoft and Intel had a virtual monopoly on the personal computer’s two most essential parts, the operating system and the microprocessor, two critical suppliers to the PC industry. While they had done little initially to develop the personal computer industry, they were successful in taking over its leadership. In the process, they extracted all the profit from the personal computer industry, resulting in becoming two of the most valuable companies in the world.
    As the personal computer began to dominate the computer industry, Andy Grove, then CEO of Intel, made an important observation. The computer industry had moved from a vertical organization in which computer companies developed and manufactured all the key elements of their computer products to a horizontal industry where there were different leaders amongst all the crucial components such as discs, software, semiconductor components, and even distribution. Andy thought not only was this very profound but that it was more or less a permeant shift. He describes this in his book Only the Paranoid Survive (page 44).


    Not only was Andy convinced, but most of us in senior roles in the computer industry agreed with him. By that time, the only significant vertical manufacture of personal computers was Apple with the Macintosh. Apple had a tiny market share made up of devoted users mainly from the creative community. Most thought that Apple would eventually fail, and it almost did. But it would turn out to be Apple and Steve Jobs that would once again turn the computer industry on its side, returning to the vertical model. Looking at the world of mobile devices that dominate computing, we can see the leading player, Apple, is vertical although out-sourcing its manufacturing and some key components, but they are proprietary. In the Android world, the operating system is an independent component similar, but otherwise, major suppliers like Samsung operate similarly to Apple. So I take it to be that the development of the personal computer industry was not a given. IBM with the PS/2 and OS/2 tried to return to the vertical model, but it was too late.


    What was inevitable with that computers would be designed for the individual? Yes! Changes in technology such as semiconductors, Winchester disks, and networking would have driven the computer industry to create computers that would have looked a lot like the IBM PC and the computers that followed. More and more people were getting exposed to using computers. But it would have also been likely that the key computer companies like IBM and Digital Equipment Corporation could have developed proprietary systems much in the way that companies like Sun and Silicon Graphics did for graphic workstations.


    In 1980, when I proposed developing the Professional Series Computer to the Operations Committee at Digital Equipment, I had knowledge workers in mind, including individuals working in small offices. Rather than break with the computer systems that Digital was offering, I saw the Professional as an extension to the ecosystem that Digital had already developed. I wanted to develop a computer designed for the individual but compatible with the company’s other computer offerings, able to run the same application and utilize the same communication network. Even though the Professional was based on the PDP-11’s 16-bit architecture, we came up with a way to run application software on it which was designed for the VAX line of computers. Years later, I would discuss this with Don Estridge, the person generally credited with the development of the IBM PC. He told me that the Professional was the personal computer that had concerned him. But he had nothing to worry about as the Professional would end up a failure.


    There were many reasons for the failure of the Professional, and I am responsible for a number of them. But the foremost reason was the advent of the IBM PC which would upend the entire computer industry, including IBM. In addition, the arrival of the IBM PC provided an opportunity for other groups within Digital to challenge my strategy. In particular, one of the company’s product lines proposes a practice similar but not fully compatible to the IBM PC, The Rainbow. Another, the Word-processor Product Line, run by Ken’s brother Stan, decided to build a dedicated Word processor, DecMate II. Ken Olsen, the CEO, loved internal completion. He was delighted to have three products and their teams going after each other. Like most new products, there were some serious flaws with the initial implementation of the Professional. It was way over-engineered, causing it to take too long to develop. It had performance problems and was too expensive. But had the IBM PC not come along when it did and had Ken Olsen not decided to end up with three competing products (He said “Let the market decided” and I said “they did. They bought IBM’s product”, I think the Profession could have been successful.

  • Fun interview that may get me into trouble

    Over the last month, I have been doing many interviews to help promote my book, The Flight of a Wild Duck, published on September 1st.  One of my favorite interviews was . It is long but lots of fun. Most of my interviews have been enjoyable, but this one has a super amount of energy due to the host’s personality, Christopher Lochhead.  I do hope you enjoy listening to it here.

    However, I said one thing that would probably get me into “trouble” with some people. I said that I did not think that Andy Grove was a great CEO. I explained why in detail. First of all, I believe that Andy Grove was a great man. I had the fortune to know him rather well. He was remarkable in so many ways. So much has been written about Andy, the man, so that I won’t repeat that here. But Andy failed at succession. I think that several attributes go into making a great CEO, and that is one. He should never have supported Craig Barrett as his successor. Barrett was a great Chief Operating Office but was not the strategic thinker and leader needed as Intel was about to enter into a significant discontinuity driven by the internet. Andy was one of the most outstanding CEOs when it came to execution, but he struggled with a vision about the future. My book goes into that in detail. Yes, under Andy’s watch, Intel grew to become one of the most valuable companies in the world, but then looked at what happened in the next twenty years.  

    I also said that I thought Jeff Bezos and Steve Jobs were great CEOs and gave my reasons. But I also said that I did not care for them as people. I wish there were a high correlation between being a great person and a great CEO, but it does not appear that there is.  

    Andy is so revered that I expect to get a lot of flack for what I said about him. But knowing Andy, I think if he was around now and could look back, he might agree with me.

  • Rereading Only the Paranoid Survive

    Only the paranoid survive

    At the suggestion of my friend, David Yoffie, I reread Andy Grove’s book, Only the Paranoid Survive, written in 1996 and I am glad I did. David, who is a professor at Harvard’s business school, served on the Intel Board of directors for almost 29 years leaving in May of 2018. David visits Israel several times a year, and we often have a meal together. Last week we went for a long walk and a great lunch. I took the opportunity to discuss the development of my book with him as he has written several successful books himself. He and I often talk about Intel as you might expect and we especially discuss Andy Grove who had a profound effect on both of our careers.

    I have a bit of “history” with Only the Paranoid Survive. Harriet Rubin, the head of the Currency Imprint at the publisher, Doubleday and I were in a long term relationship when Harriet approached Andy with the idea of writing a business book. As Andy says in the preface of the book, “I had no intention of writing a book on this subject until Harriet Rubin, from Double day, sought me out and convinced me that I should do so.” Around this time, Harriet would join me for the Sun Valley Allen & Co conference. Andy attended as well so I would guess that this helped Harriet, who published some of the most impactful business books of the 90s, develop her pitch to Andy. I did not discuss the book much with Harriet during the time Andy was writing it. I felt a bit awkward about the situation although Andy did not seem to mind. One of the people that had been working on my team, Rob Siegel, spent about six months, I think, helping Andy with research on the book.

    After the book was published, Andy gave me a copy of the book on which he wrote, “I have dedicated chapter 7 to you”. The title of Chapter is “Let Chaos Reign.” It says a bit about how Andy viewed me.

    In re-reading the book, I was struck with the strategic insight that Andy showed. It is even more impactful when you think of his journey as a Jew from Nazi dominated, and then Russian dominated Hungry, to the USA and later from chemical engineer to the CEO of one of the most successful companies in the world, Intel.

    This book is full of insight and wisdom. I plan to build upon many of the things that Andy describes. Rereading the book also made me wish that I was able to have a better relationship with Andy. I could have learned more from him and frankly, he could have learned more from me. I have grown to think that Andy had more respect for my abilities and thinking that I believed at that time. Someone very close to Andy recently told me “He loved you. Said you always needed a wild duck like Avram. A nonlinear thinker that stirred up the others. He felt your presence was quite critical and when you left there really was never anyone like you. I always try to have an “Avram” on my team. Always. They disrupt and create. Don’t confuse that he didn’t always listen with how he felt”.

    I invite you to read Only the Paranoid Survive.

  • The Andy Grove I knew

    Yesterday, Andy Grove died. Frankly, I am still processing that.  I first met Andy in 1982 at the PC Forum.  Andy was 46 and I was 37.  The next time I met him was in 1984 when I was interviewing for a position at Intel. I would be reporting to Les Vadasz, who, like Andy, was a Jewish Hungarian refugee.  I interviewed with both Gordon Moore and Andy Grove, to whom, Les reported. I must have done something right that day, because, during the interview, Andy picked up the phone and told Les that he should hire me. I can’t say that I had many other days like during the times I worked with Andy.  He and I  had a kind of rocky relationship over the fifteen years that I worked at Intel.  I learned a great deal from Andy.  Sadly, I don’t think the opposite was true.  Andy never really got me, but on the other hand, was very supportive and respectful in his own way.  When I thought about leaving Intel out of frustration about three years after joining, Andy talked me into staying.  I will never forget him telling me that I should stay “because we accept you.”  I think he meant that while I was weird, that was ok because I shared many common values with Andy and Intel including the ability to get things done.  When Les proposed that I be made a Corporate Vice President (one of just 33 officers), Andy supported Les and sure enough, I became an officer of one of the most successful companies in the world. 

    I felt that Intel and in particular, Andy, was too insular and I tried to introduce Andy to people from other industries. I was able to get him to join me at the Allen&Co conferences in Sun Valley.  He loved going, I think.  But then when I brought Steven Spielberg to visit Intel and asked him to join the meeting, he called me a “star fucker.”  I got to bust him later when I saw a photo that was taken that day with Andy and Spielberg, taped to the refrigerator in the kitchen of his home. Over the years, I did see that Andy became more outgoing and more subtle.  But I was often frustrated, I guess with myself because I did not have the impact on this thinking that I wanted.  In particular, I failed to get him to understand the Internet was not just a way of connecting PCs together. 

    Once, I sent him a memo calling him the Mick Jagger of Technology.  I said that he, like an aging rock star, just wanted to make today last.  But that incidentally also demonstrated something special about Andy.  He could dish it out (and oh how he could dish it out) but he could also take it. I probably went up a notch in his thinking because of that email although I am sure he did not take is the message to heart.

    Andy was there for me when I got Prostate Cancer. He was in the class a year ahead of me. He was very concerned about my health.  But in the typical Andy fashion, he became fanatic about his diet and expected me to eat the same way.  I ended up hiding out in the cafeteria when I eat cottage cheese.

    Andy stepped down as CEO in 1998 but stayed Chair of the Board for a while. I left Intel in 1999.  I didn’t see him very often.  Les Vadasz stayed close to him and through Les and other friends, I learned about Andy’s Parkinson disease. It really upset me to think that Andy would be robbed of his health during the years when he could have turned his talents to other pursuits.  Of course, he did his best during those difficult years but how painful it must have been for him both physically and mentally and how sad and difficult for his wife, Eva.   Andy taught me one more lesson which is the importance of moving on. 

    The last time I saw Andy was in 2013 at the 45th anniversary of Intel. He was there with his wonderful wife, Eve.  I went up to him and said, “give me a fucking hug”.  He did and I am very glad that was my last experience of Andy.

  • Andy Grove: Only half right

    Andy wrote an article about job creation or the lack of said  recently.   I think Andy is totally right that  high tech start ups are not the answer to domestic job creation.  Large scale job creation for high tech can only happen if there is local manufacturing and I do not see that in the world we now live in.  Even software and support services will move abroad if companies are successful although to a smaller degree.  It is natural for companies to seek lower cost and we as consumers benefit from that.   I was shocked that Andy, felt that taxing products that contained outside labor was a good idea.  It is an awful idea in my opinion.  First of all that is a tax on US Citizens in that we would have to pay more for products to that we could support American workers but more importantly it could destroy the world economy (if it is not already destroyed by the foolish actions of various governments).  And we need to see countries like China, India etc develop for a number of reasons.  On the other hand, we should push for fair valuation of Chinese currency, work standards etc.

    There are a number of actions our pathetic government (and as always I mean all three branches and both major political parties) could take to increase domestic employment such a major creation of green jobs in both energy creation and most importantly in the short term, energy conservation, the basic problem is that there will not be enough jobs for Americans and we need to build a new model to deal with that.  We were once a farming nation.  Now we have twice as many people in prison than working on farms.  We need to convert our major military industrial complex to creating such things as high speed rail. We need to make our education system work so that we create even more knowledge workers.  There should be two years of nation service young people.    These are just a few ideas.  But most importantly, employment will stay high and we will get use to it.  It is awful of course for those that are unemployed but we will not help them with taxes on imports.