Array
(
    [content] => 
    [params] => Array
        (
            [0] => /forum/index.php?threads/pat-gelsinger-interview-economic-club-washington-dc.15219/
        )

    [addOns] => Array
        (
            [DL6/MLTP] => 13
            [Hampel/TimeZoneDebug] => 1000070
            [SV/ChangePostDate] => 2010200
            [SemiWiki/Newsletter] => 1000010
            [SemiWiki/WPMenu] => 1000010
            [SemiWiki/XPressExtend] => 1000010
            [ThemeHouse/XLink] => 1000970
            [ThemeHouse/XPress] => 1010570
            [XF] => 2021370
            [XFI] => 1050270
        )

    [wordpress] => /var/www/html
)

Pat Gelsinger Interview Economic Club Washington DC

Daniel, thanks this is very interesting and I learnt a lot from this. Huge respect for Pat for getting to the head of Intel from such a junior starting position and for what he's done in his career.

He's certainly also making some really big calls and not hiding from the issues. Again welcome and good.

However, I am a little concerned that with his natural reverence for the Intel founders (he mentions Moore and Grove) and his admission that he modelled himself on them whether he is enough of an "outsider" to effect the sort of change that may be needed. My sense - and I think that Pat himself has said this - is that this is cultural change as well as business and technical change. Is bringing back some of the old crew - capable as they may be - the right way to do this ?

Changing on all three fronts at the same time is a really tall order and it may take more than one person. I'd like to see someone else coming forward to give me confidence that this is a real team effort from the top. Where are the other leaders pushing through these changes and why aren't they speaking up with him too ?

As I think I may have said before, Pat seems to be taking on several different roles at the same time and risks coming across as both "good cop" and "bad cop" at the same time when these roles are probably best done by different people. He could delegate that.
 
To give Pat due credit. He’s smart, has legendary work ethic, understands the industry, and most importantly is willing to pivot. And I think ultimately that’s what he’s going to need to do - pivot. He did it with VMWare when the data center business was disrupted by cloud and he realized that fighting Amazon was a losing battle. And he’s going to do it again because fighting TSMC is also losing battle.
 
Being an optimist, I find it hard to be a downer and a pessimist, but I’m also a realist and a fairly good judge of character. I don’t know Pat Gelsinger personally, but what I’ve witnessed from his personal behavior, intertwined with third-party assessments, I believe Pat is an opportunist, who by nature finds a crack and exploits it, and often, likely, recklessly.

I’m a successful engineer, so I understand the drive to solve, and successfully finding solutions requires finding and analyzing cracks, but I have seen many scientists without true intelligence, and I believe Pat is one of the latter.

I sincerely hope to be proven wrong. Time will tell.
 
Pat's legacy will be tightly coupled with "Intel Accelerated". Announced July 2021, “Intel Accelerated” is a very aggressive roadmap for Intel to regain process leadership:

2H-2022 – 4nm (formerly 7nm) – 20% performance per watt improvement, first EUV use at Intel.

2H-2023 – 3nm – 18% performance improvement, denser libraries (some density increase) and more EUV use.

1H-2024 – 20A (2nm) – RibbonFET (first generation Intel GAA), a type of horizontal nanosheet.

1H-2025 – 18A (1.8nm) – RibbonFET (second generation Intel GAA).

If Pat can deliver on this promise he will be in the Intel ranks of Andy and Gordon no matter what the foundries do. If the foundries stumble on GAA there will be no stopping Pat, my opinion.
 
Pat's legacy will be tightly coupled with "Intel Accelerated". Announced July 2021, “Intel Accelerated” is a very aggressive roadmap for Intel to regain process leadership:

2H-2022 – 4nm (formerly 7nm) – 20% performance per watt improvement, first EUV use at Intel.

2H-2023 – 3nm – 18% performance improvement, denser libraries (some density increase) and more EUV use.

1H-2024 – 20A (2nm) – RibbonFET (first generation Intel GAA), a type of horizontal nanosheet.

1H-2025 – 18A (1.8nm) – RibbonFET (second generation Intel GAA).

If Pat can deliver on this promise he will be in the Intel ranks of Andy and Gordon no matter what the foundries do. If the foundries stumble on GAA there will be no stopping Pat, my opinion.
Intel should be 5-9 months behind tsmc. Tsmc uses nova equipment and has the technical people this is a simple fix that went on to long.

People don't talk about the design productivity gap enough. It hurts intel.

Pat gelsinger is no Andy grove. Grove was more than a ceo.
 
Pat's legacy will be tightly coupled with "Intel Accelerated". Announced July 2021, “Intel Accelerated” is a very aggressive roadmap for Intel to regain process leadership:

2H-2022 – 4nm (formerly 7nm) – 20% performance per watt improvement, first EUV use at Intel.

2H-2023 – 3nm – 18% performance improvement, denser libraries (some density increase) and more EUV use.

1H-2024 – 20A (2nm) – RibbonFET (first generation Intel GAA), a type of horizontal nanosheet.

1H-2025 – 18A (1.8nm) – RibbonFET (second generation Intel GAA).

If Pat can deliver on this promise he will be in the Intel ranks of Andy and Gordon no matter what the foundries do. If the foundries stumble on GAA there will be no stopping Pat, my opinion.
They're going to need a lot of good luck achieving each one of these 4 milestones because a node improvement per year is unprecedented even by the best of the best in the industry. I'd say pretty soon we'll see them push back the timelines because it's flat out unrealistic and even more so by a company who has been a laggard in this area. Frankly I'd be impressed if they can hit 3nm by 2025.
 
Pat's legacy will be tightly coupled with "Intel Accelerated". Announced July 2021, “Intel Accelerated” is a very aggressive roadmap for Intel to regain process leadership:

2H-2022 – 4nm (formerly 7nm) – 20% performance per watt improvement, first EUV use at Intel.

2H-2023 – 3nm – 18% performance improvement, denser libraries (some density increase) and more EUV use.

1H-2024 – 20A (2nm) – RibbonFET (first generation Intel GAA), a type of horizontal nanosheet.

1H-2025 – 18A (1.8nm) – RibbonFET (second generation Intel GAA).

If Pat can deliver on this promise he will be in the Intel ranks of Andy and Gordon no matter what the foundries do. If the foundries stumble on GAA there will be no stopping Pat, my opinion.

Can it be possible that Intel's real internal target is Intel 20A in 2024 or 18A in 2025? By then Intel's new fabs will be ready and they will finally procure just enough number of EUV machines to have a meaningful production capacity.

Before 2024 or 2025, Intel 4 and Intel 3 will have much lower production capacity and product availability. Basically they will be the "Me Too" nodes to show the world that Intel is on track to regain the technology leadership.

During this transition time, Intel will try to use various ways to fill the gaps and maintain the market share as much as they can. TSMC's capability is probably one of the work around to help Intel to lessen the stress.

Intel may also cut the price or do something equivalent to keep the customer. I suspect Intel may be doing price cutting on selected products already. It may partially explain while most major semiconductor companies are getting 20 - 50 percent growth in 2021, Intel is forecasted to have a negative 1% growth in 2021. As we all know many semiconductor products are in short supply across multiple industries globally. How can Intel manage to get such miserable and strange 2021 results?

I'm just guessing...
 
Last edited:
During this interview, Pat Gelsinger stressed the importance of diversity, STEM/STEAM education, and hiring more women at Intel. I'd like to share three videos from Society for Science, the organizer of the renown Regeneron Science Talent Search (for high school seniors, celebrating 80th anniversary) and Broadcom MASTERS science award (for middle school). It's really an enjoyment to watch these children's passion in science and their enthusiasm for a better world. The diversity nature of these two awards can be seen vividly from those participants.





In 2015 Brian Krzanich, then Intel CEO, decided to stop their 18-year sponsorship to the Intel Science Talent Search. Later Regeneron Pharmaceuticals took over the sponsorship and the award was renamed to Regeneron Science Talent Search. How do I believe Intel is serious about diversity, gender fairness, and STEM education when Intel thought $6 million (a tiny portion of Intel's annual spending) yearly contribution to STS was not a smart spending?

 
Last edited:
Back
Top