Intel CEO Paul Otellini to Retire in May(newsroom.intel.com)
newsroom.intel.com
Intel CEO Paul Otellini to Retire in May
http://newsroom.intel.com/community/intel_newsroom/blog/2012/11/19/intel-ceo-paul-otellini-to-retire-in-may?cid=rss-90004-c1-278408
37 comments
The first non-tech CEO is usually a strong sign that a company has grown from a startup with a single product focus to a long-term viable business with a diversified product portfolio. (Think GE, AT&T, Boeing etc.)
For people in the HN community, this is mostly horrific because it represents everything they avoid. However, for almost everyone else out there that is working for the purpose of simply providing for their families, the stability and strength of a diversified company is typically a major positive.
The thing that is always missing from the HN discussions about how bad MBAs are for a company etc. is that at the extremes of business, business becomes highly technical and requires someone who is technically proficient at business.
Knowing how to create the next high end transistor technology which maintains Moore's law is amazing, but that person is highly unlikely to keep a company of Intel's size successful.
For people in the HN community, this is mostly horrific because it represents everything they avoid. However, for almost everyone else out there that is working for the purpose of simply providing for their families, the stability and strength of a diversified company is typically a major positive.
The thing that is always missing from the HN discussions about how bad MBAs are for a company etc. is that at the extremes of business, business becomes highly technical and requires someone who is technically proficient at business.
Knowing how to create the next high end transistor technology which maintains Moore's law is amazing, but that person is highly unlikely to keep a company of Intel's size successful.
that person is highly unlikely to keep a company of Intel's size successful.
Perhaps so... particularly in the moment. To me, it's more a question of whether that CEO is a person who once would have known how to create the next high end transistor (or whatever technology innovation is central to the business) and is still able to have that conversation as an engineer/scientist.
For instance, I doubt Andy Grove, at the end of his tenure, would have been able to do that kind of tech work in the moment, but he had a PhD in engineering and had a very deep knowledge of the technology as well as the business. He did make some very typical-of-engineers mistakes when intel decided to raise its consumer profile. And I'm not trying to bring up a debate around Andy Grove's management style (though of course, people are free to comment what they like). But it's hard to say he didn't keep the company successful, even at a very large scale.
Otellini was a successful CEO, but (I actually don't know this for sure) did he ever have the ability to do the core tech work behind intel's product?
I suppose that in this era of high profile non-tech CEO flameouts (HP, Yahoo), Otellini may actually be a counter-example that a non-technical CEO can still be quite successful.
Perhaps so... particularly in the moment. To me, it's more a question of whether that CEO is a person who once would have known how to create the next high end transistor (or whatever technology innovation is central to the business) and is still able to have that conversation as an engineer/scientist.
For instance, I doubt Andy Grove, at the end of his tenure, would have been able to do that kind of tech work in the moment, but he had a PhD in engineering and had a very deep knowledge of the technology as well as the business. He did make some very typical-of-engineers mistakes when intel decided to raise its consumer profile. And I'm not trying to bring up a debate around Andy Grove's management style (though of course, people are free to comment what they like). But it's hard to say he didn't keep the company successful, even at a very large scale.
Otellini was a successful CEO, but (I actually don't know this for sure) did he ever have the ability to do the core tech work behind intel's product?
I suppose that in this era of high profile non-tech CEO flameouts (HP, Yahoo), Otellini may actually be a counter-example that a non-technical CEO can still be quite successful.
>For instance, I doubt Andy Grove, at the end of his tenure, would have been able to do that kind of tech work in the moment, but he had a PhD in engineering and had a very deep knowledge of the technology as well as the business. He did make some very typical-of-engineers mistakes when intel decided to raise its consumer profile.
Grove is a great example of a successful technical CEO. I would never say Otellini was a failure, but Intel flourished under Grove in a way that it didn't under Otellini.
Grove is a great example of a successful technical CEO. I would never say Otellini was a failure, but Intel flourished under Grove in a way that it didn't under Otellini.
Yes, I agree, to say Grove "kept the company successful" (referring to once it became a huge company), is to understate how much it flourished.
What a legacy -- he retires the same day that Intel wins. AMD just ceded the high margin CPU business to concentrate on the volume CPU business. https://news.ycombinator.com/item?id=4804602
Otellini's successor will be playing a different game, fighting ARM rather than AMD.
Otellini's successor will be playing a different game, fighting ARM rather than AMD.
>Otellini's successor will be playing a different game, fighting ARM rather than AMD.
Contra the ARM fans who seem to be dominant on HN, I think Intel is well positioned to win that battle. Cutting-edge fab capacity is a pretty big advantage, even if they're using it mostly for high-power CPUs at the moment.
Contra the ARM fans who seem to be dominant on HN, I think Intel is well positioned to win that battle. Cutting-edge fab capacity is a pretty big advantage, even if they're using it mostly for high-power CPUs at the moment.
Which is the game Intel should have been playing for the last N years.
Reports are that the Razr i based on an Intel processor is competitive -- similar performance & power usage to an S4 with twice the number of cores. The pre-game warmup is complete. Now the game starts.
Right, they're playing the game and have competetive parts. What the ARM-is-the-future people fail to recognize is that the x86 world isn't remotely dead. Relative to consumer SoCs, Intel makes shockingly high margins on its desktop and server silicon. It simply doesn't make sense for them to be making high volume embedded chips unless they have excess fab capacity (which they don't).
The Medfield chips are performance-competetive with existing parts, though they aren't stand-outs. But even if, say, they had a mythical chip that would get lets-just-say-for-argument 2x the performance of Tegra 3 or A6 at 50% of the power draw, they would still not be manufacturing it in volume, for the simple reason that Ivy Bridge cores make a whole lot more money for them than a Tegra-killer would.
So from an investor's perspective, Intel is doing exactly the right thing. Stay in the game, don't cede the market. Make sure your products are competetive, but don't walk away from easy cash either.
Long term, their success will be driven by the value of their SoC products as the desktop world dissolves. But current market penetration isn't a good indicator of what the likelihood of that value will be.
The Medfield chips are performance-competetive with existing parts, though they aren't stand-outs. But even if, say, they had a mythical chip that would get lets-just-say-for-argument 2x the performance of Tegra 3 or A6 at 50% of the power draw, they would still not be manufacturing it in volume, for the simple reason that Ivy Bridge cores make a whole lot more money for them than a Tegra-killer would.
So from an investor's perspective, Intel is doing exactly the right thing. Stay in the game, don't cede the market. Make sure your products are competetive, but don't walk away from easy cash either.
Long term, their success will be driven by the value of their SoC products as the desktop world dissolves. But current market penetration isn't a good indicator of what the likelihood of that value will be.
> It simply doesn't make sense for them to be making high volume embedded chips unless they have excess fab capacity (which they don't).
Well they desperately want to do it anyway. Anyway by your very reasoning, they won't do high volume chips (you added embedded but it does not matter, you just can't reach the same volume as embedded chips with such higher prices). Whether they not reach that for a reason or another does not matter much; but the consequences of that fail could be hard. The desktop and server world is not going away, the x86 world also is here to stay, but there is no hard limit to the changes of market repartition that can happen even there. IBM is still very powerful with their mainframes, and Intel could follow that path for (a part of) desktops and servers, but the more they allow the ARM world to develop, the more their share is at risk on their own land in the long term.
The problem is that SoC offering not only matters now, but it even has mattered for several years, and Intel had no offering, sill has no serious one, and is probably not even going to have no serious one for several other years. For a company with such resources, this is crazy. And when you take a look at the press communicate to try to understand what they think that matter, you do not expect big changes: "operations and the cost structure" i don't know enough to judge so i give it the benefit of the doubt, "breakthrough innovations" are real important stuff, so good point here; but: - "Reinvented the PC with Ultrabook™ devices" <- is this a joke??? - "Greatly expanded business partnerships and made strategic acquisitions that expanded Intel’s presence in security, software and mobile communications" <- ok they are talking about Intel's McAfee and other stuffs that makes no sense here - "Delivered the first smartphones and tablets for sale with Intel inside" <- okkkkk, so? huge fail here, and they do not even try to pretend they matter in this field, just that they exist, because they know they won't be forgiven if they don't. That's what we are talking about. - "Grew the vast network of cloud-based computing built on Intel products" <- not driven by them. They could as well have talked about the internet and we servers, but the fashion word of the day is cloud, so they talked about that.
Intel, desktops and servers, are here to stay, we agree on that. But their market share is at risk, because these markets are not isolated island. Intel won't be able to increase their prices too much, so if their market decrease too much, do the maths.
Well they desperately want to do it anyway. Anyway by your very reasoning, they won't do high volume chips (you added embedded but it does not matter, you just can't reach the same volume as embedded chips with such higher prices). Whether they not reach that for a reason or another does not matter much; but the consequences of that fail could be hard. The desktop and server world is not going away, the x86 world also is here to stay, but there is no hard limit to the changes of market repartition that can happen even there. IBM is still very powerful with their mainframes, and Intel could follow that path for (a part of) desktops and servers, but the more they allow the ARM world to develop, the more their share is at risk on their own land in the long term.
The problem is that SoC offering not only matters now, but it even has mattered for several years, and Intel had no offering, sill has no serious one, and is probably not even going to have no serious one for several other years. For a company with such resources, this is crazy. And when you take a look at the press communicate to try to understand what they think that matter, you do not expect big changes: "operations and the cost structure" i don't know enough to judge so i give it the benefit of the doubt, "breakthrough innovations" are real important stuff, so good point here; but: - "Reinvented the PC with Ultrabook™ devices" <- is this a joke??? - "Greatly expanded business partnerships and made strategic acquisitions that expanded Intel’s presence in security, software and mobile communications" <- ok they are talking about Intel's McAfee and other stuffs that makes no sense here - "Delivered the first smartphones and tablets for sale with Intel inside" <- okkkkk, so? huge fail here, and they do not even try to pretend they matter in this field, just that they exist, because they know they won't be forgiven if they don't. That's what we are talking about. - "Grew the vast network of cloud-based computing built on Intel products" <- not driven by them. They could as well have talked about the internet and we servers, but the fashion word of the day is cloud, so they talked about that.
Intel, desktops and servers, are here to stay, we agree on that. But their market share is at risk, because these markets are not isolated island. Intel won't be able to increase their prices too much, so if their market decrease too much, do the maths.
The new Intel phones didn't build themselves overnight.
We can all argue that N should be higher than it actually was, but Intel's still present and accounted for.
We can all argue that N should be higher than it actually was, but Intel's still present and accounted for.
AMD has denied that report and the article now reflects this info. FTA: Updated 11/19/2012@10:15am: AMD contacted us with an official denial of the story and stated that Kaveri and the big cores are still on track.
So according to Moore's law, he should have seen things get 106528681 times more transistors at Intel!
Moore's law: Transistor count doubles every 18 months. Assuming 40 years at the company (it's a bit less) that's: 2^(40/1.5) = 67,108,864
The 62 Core Xeon Phi has 5 billion transistors. Divided by that number is: 149. That's a bit low. The Intel 8080 that came out in 1974 (when he joined the company) had 4,500 transistors. Looks like Moore's law has been slowing down a bit in recent years. Probably due to the focus on reduced energy usage.
EDIT: It works out if you replace "18 months" by "24 months". Revised value: 2^(38/2) = 524288. Impressive still.
The 62 Core Xeon Phi has 5 billion transistors. Divided by that number is: 149. That's a bit low. The Intel 8080 that came out in 1974 (when he joined the company) had 4,500 transistors. Looks like Moore's law has been slowing down a bit in recent years. Probably due to the focus on reduced energy usage.
EDIT: It works out if you replace "18 months" by "24 months". Revised value: 2^(38/2) = 524288. Impressive still.
2^(40/1.5) is 106528681, just like I posted...
Which calculator are you using? Does it use a Pentium CPU with the famous division bug perhaps? :)
Which calculator are you using? Does it use a Pentium CPU with the famous division bug perhaps? :)
Bummed to see that the financial achievements were noted before the technical ones (and that the true technical achievements were relegated to a single bullet point).
This is hard to do at scale: "* Achieved breakthrough innovations, including High-K/Metal gate and now 3-D Tri-gate transistors; and dramatic improvement in energy efficiency of Intel processors"
This is hard to do at scale: "* Achieved breakthrough innovations, including High-K/Metal gate and now 3-D Tri-gate transistors; and dramatic improvement in energy efficiency of Intel processors"
Notable absent from that list is he fired Pat Gelsinger.
Please just bring him back as CEO.
Please just bring him back as CEO.
That's a very long time to work for one company! Wish him the best.
I work for Intel, and I congratulate Paul on his tenure.
On the whole, I would say most of my engineer colleagues approve of Paul's leadership, MBA or otherwise :).
It will be interesting to see if Intel's next CEO decides to take any exciting new turns. Particularly with respect to the PC market, we live in interesting times.
On the whole, I would say most of my engineer colleagues approve of Paul's leadership, MBA or otherwise :).
It will be interesting to see if Intel's next CEO decides to take any exciting new turns. Particularly with respect to the PC market, we live in interesting times.
Nit pick: The article says he's been at Intel for "nearly 40 years". According to Wikipedia he joined the company in 1974 [1]. Still a long time though.
[1] http://en.wikipedia.org/wiki/Paul_Otellini#Employment_at_Int...
[1] http://en.wikipedia.org/wiki/Paul_Otellini#Employment_at_Int...
Is 38 not nearly 40?
My point related to the title of the HN item, which is "Intel CEO Paul Otellini retiring in May after 40 years with company". This is different to the title of the original article, which doesn't mention the duration.
A good time for him to leave before the downwards trend of Intel becomes more obvious, and people start blaming it on him. I hope he leaves the board of Google, too, because I don't like how he has influenced some of Google's decisions in the past few years, to use Intel chips instead of ARM in their new devices, and every single one of them turned out to be the wrong decision.
Which "new devices" are you referring to? As far as I know the Nexus One and Nexus 4 use Qualcomm CPUs, and the Nexus S, Galaxy Nexus, Nexus 4, and Nexus 10 all use ARM CPUs.
Edit: Thanks to those who pointed out that I'd forgotten the Chromebooks and Google TV. Given the limited extent to which Google have marketed these devices, I'm not sure whether they can be said to have failed yet. And even if they have failed, it is not clear that this was due to using Intel components.
Edit: Thanks to those who pointed out that I'd forgotten the Chromebooks and Google TV. Given the limited extent to which Google have marketed these devices, I'm not sure whether they can be said to have failed yet. And even if they have failed, it is not clear that this was due to using Intel components.
But those were successful devices. He's talking about Google devices that failed, in part because they used Intel CPU's. Devices like the original Google TV.
How do you know it failed because of Intel CPUs?
and Chromebooks.
All the unsuccessful Chromebooks and the Revue Google TV. Both of those have only recently started to pick-up when they started using ARM chips. And now Google is repeating the mistake by starting to put Intel chips in its Motorola devices.
This is also a very bad strategic mistake because if Intel succeeds in mobile, and becomes popular, who do you think it will help most? Google? No way. That situation would help Microsoft, because they can start using the x86 chips that got popular on Android to promote their x86 tablets and hybrids, where Microsoft has a real strength because of legacy apps.
So Google is now pushing Intel chips into the market, and gaining who knows what, because I don't see any benefit for them, but definitely helping Microsoft in the long term. If Intel survives, Microsoft survives. You'd think Google would realize that and doesn't want that.
Google has absolutely no need for Intel in the future, even for their data centers. Microsoft's future pretty much depends on Intel and x86 chips. I don't understand why Google isn't taking full advantage of this weakness of their competitor, and instead it's trying to put Intel back in the game.
This is also a very bad strategic mistake because if Intel succeeds in mobile, and becomes popular, who do you think it will help most? Google? No way. That situation would help Microsoft, because they can start using the x86 chips that got popular on Android to promote their x86 tablets and hybrids, where Microsoft has a real strength because of legacy apps.
So Google is now pushing Intel chips into the market, and gaining who knows what, because I don't see any benefit for them, but definitely helping Microsoft in the long term. If Intel survives, Microsoft survives. You'd think Google would realize that and doesn't want that.
Google has absolutely no need for Intel in the future, even for their data centers. Microsoft's future pretty much depends on Intel and x86 chips. I don't understand why Google isn't taking full advantage of this weakness of their competitor, and instead it's trying to put Intel back in the game.
Trust me when I say that the lack of success of Chromebooks and Google TV have very little to do with Intel.
With that said, Intel has been way too slow in its race against ARM, but saying that this has impacted Google I don't find credible.
With that said, Intel has been way too slow in its race against ARM, but saying that this has impacted Google I don't find credible.
The ARM Chromebook costs less than x86 Chromebooks while offering more-or-less the same user experience and it's much more popular in sales, according to Amazon.
So, Intel is/was a part of the problem at least for Chromebooks. (I don't know anything about Google TV)
So, Intel is/was a part of the problem at least for Chromebooks. (I don't know anything about Google TV)
There's also motorola razr-i (europe only).
This is the most biased and unscientific statement. Where the hell is the downvote button? How does using Intel chips make Google products unsuccessful? Apple has been using Intel chips for several years (although they want to move to ARM soon), and Apple products are/were very successful. The scientific explaination to the failure of ChroomBook is a revolutionary idea. People want to access their data even on the train.What makes you think a cloud-based OS will win? Duh. Wifi sucks. Home Internet sucks too. TV? Apple TV sucks too, why would Google TV be any different? Netflix works with TV vendors and embedded webapp into TV. That's included in the TV price. More useful than an extra box.
Actually, I think "first non-technical CEO" probably deserves a notch on a technology company's time line. It doesn't mean that the company won't succeed, but it is a sign that it has become a different type of company (this may be more a reflection of this change than a cause).
But unlike some other high profile flame-outs, it sounds like Otellini was a success at the helm.