Intel CEO Bob Swan is stepping down from the position on February 15th, the company has announced. He will be replaced by VMware CEO Pat Gelsinger. Swan was named Intel’s permanent CEO two years ago in January 2019. He initially took on the role on an interim basis in June 2018 following the resignation of Intel’s previous CEO Brian Krzanich.
They need a Lisa Su.
Even Lisa Su wouldn’t be able to do a thing to steer the ship around if she wants to farm out chip production to TSMC but several factions within Intel struggle to keep Intel designs exclusive to Intel’s foundries. Or vice-versa (you never know). Intel’s current position is not an easy position for any company to be in, since Intel’s foundries have traditionally been the company’s biggest strength and a major differentiator from AMD until very recently. There is fear inside Intel that if the company decides to take a bite of that sweet TSMC silicon, management will gradually divest the foundries to enjoy short-term gains, which could possibly mean layoffs (and resulting managerial friction) and a loss of the company’s main differentiator in the long-term.
Intel hasn’t faced such a pivotal moment ever since they walked away from the memory business. With the difference that the current situation is not as clear-cut as it was back then. You see, Andy Grove’s decision was easy: Memory was becoming increasingly commodified, so they walked away. Have the foundries become commoditized? And if yes, why are there only two companies on cutting-edge nodes?
Meanwhile Intel is stuck with “pale male and stale” cut-and-paste executive reshuffles and purple faced internal factional warfare on an elevator to hell. Going down.
congrats, you managed to be rasist and sexists in 3 words.
Oh, so only right wingers are allowed to tell jokes? ‘Scuuuuuuuuuuuse me for breathing! And the reality?
I genuinely am not interested in Intel or having a blow by blow account of their goings on and spent all of ten seconds thinking about that comment. I figured nobody was going to read the page long comment I spent 20 minutes drafting on geo-politics and regional trade and cooperation, or engineering production and quality standards, or how mindsets and values, and different histories may play into the situation which pretty much everyone in the industry all over is involved with. I also spent at least this long drafting and emailing a client who wanted help getting himself out of a mess, The client keeps me in business. Intel doesn’t. Easy decision.
Shouldn’t own foundries give Intel a fundamental cost advantage? Shouldn’t they still have a lot of room to undercut AMD offerings on price?
They would need to introduce some new line of products to do that (not to kill the I7 brand) but that’s a marketing job.
It should give them a cost advantage, and it did in the past when they were the leader in miniaturizing the production process. But Intel has made a couple of wrong bets (xScale, P4, Itanium and more recently EUV) that put them on a dead end road where they have hardly improved their performance and performance per Watt in the last 5 years.
ARM and AMD have entirely caught up with Intel and then surpassed them in both real world metrics and production processes. And when you use the best production process you can produce better silicon for lower cost, exactly what Intel was always best at.
And now that the world has moved from CPU to SOC for almost everything and Intel is only investing in CPU(+GPU) but not the entire SOC there doesn’t really seem a way back for Intel
Of course you should never count out Intel. They have very deep pockets, great engineers and a whole bunch of legacy and patents and they have survived a similar situation 15 years ago (Core Duo) but this time no such miracle solution seems to be available and the competition is coming from all directions
Cancelling xScale was a strategic mistake, but not a deadly one. Itanium wasn’t as much a wrong bet but a “what if” plan to lock people into a single-vendor architecture (fortunately foiled by AMD after they expanded x86 to 64-bits). Intel could cancel Itanium any moment and not suffer a loss, and in fact they did. The P4 architecture was summarily replaced by a derivative of the Pentium M architecture, so the pain was short in duration.
Intel’s floundering with regards to their manufacturing process *is* a potentially deadly mistake.
kurkosdr,
I agree. They may be able to recover, but they’ve made some major missteps. I really hope we don’t see more consolidation in the FAB market. Everyone at the high end being dependent a single fabricator is dangerous.
Isn’t Pat Gelsinger “exactly that”? A very accomplished engineer with leadership skills and a good overview of the tech landscape?
* Architected the 486
* Co-wrote a programming the CPU book
* Lead the architecture department for a long time (CTO)
* CEO at VMWARE/EMC
My main worry about him is that he seems to be very traditional, both in a personal and professional capacity. I think he would be great at focussing Intel to regain/improve in the datacenter, but not so great at transforming Intel into something the world is going to appreciate the next decade
Dirk Meyer was a very accomplished engineer with multiple successful CPU designs but a less than successful CEO because of the designs that had been developed under his predecessor. Lisa Su was in a good position becoming CEO as the Zen design developed during Rory Reed’s time was nearing viability. Not canceling a chip design when the Engineering Samples look impressive is not a definitive CEO challenge. Lisa Su’s test will be in a few years when the Zen design starts hitting its bottlenecks.
Pat Gelsinger will reap the rewards or blame for what Intel has been developing for the last few years. Given it takes 5 years for a chip to go from conception to release, Gelsinger will probably be gone from Intel before the results of Gelsinger’s strategy will be known.
So Dirk Meyer is a former CEO hampered by legacy bad designs in the pipeline and Lisa Su is a brilliant CEO just because she got lucky? Am I reading this right? Maybe Dirk meyer and his predecessor were two duds in a row and AMD has had two very good CEO’s in a row?
I almost wish I had posted my earlier draft but it was a bit waffly. What I think is really important is Lisa Su has created energy and diversity and teamwork, I suspect, both within and without AMD. This is partly because she has skill and partly because of her culture.
Timothy Garten-Ash writes in today’s Guardian about the winners and losers culture and how this is impacting European and US trade and poltiical relationships. He also examines the dynamics of the EU and its member states. It’s really important to pause at this point and consider European do what Europeans do, Americans do what Americans do, and so on for Russia, China, the Bric states and general Far-East etcetera. Then there is quality standards and internal markets and, yes, outsourcing and business networks and relationships and attitudes.
If you follow through there’s also things like the idea of fabrication “nodes” has broken down and “nodes” and now simply a marketing term to cover not just greater miniaturization but varietys of technologies and techniques to achive a fabrication goal. That depends on energy and diversity and good communication not just on a personal and business level but also the interoperating of different skills. Looking from the outside and with no special knowledge this seems to be Intels major failing. Intel got too used to its monopoly and monoculture and taking shortcuts. Would I want to work for Intel? I’m sure it’s no loss to Intel but this would be a no for lots of reasons.
It’s the same with the Raspberry PI. It’s not just because it’s affordable and accessible and it’s not just because it has an ARM cpu (i.e. not Intel) but what it represents. There’s a lot of “soft” factors behind it not just technology. Who gets excited about Intel NUC’s? I can’t think of anyone.
I’m not going to badmouth Pat Gelsinger but he seems like more the past to me and getting too long in the tooth to change. I really think if Gelsinger was the magic bullet Intel had been waiting for we would already know.
HollyB,
I really don’t know much about any of these CEOs, so don’t take anything I say as criticism of any of them personally. Now I actually agree with krebizfan’s point generally. The costs and benefits of investments and decisions often do lag those responsible for making them. It takes some time to change momentum. Meanwhile the public often likes to attribute every win and loss to whoever is in charge since it happens “on their watch” even if it was a culmination of decisions several years in the making. Anyways, I looked it up Lisa Su’s history and she’s been president for ~6 years. So personally I think that’s long enough to establish her credibility.
It’s also true AMD benefited significantly from awful missteps by it’s biggest rival intel. This was very fortunate for AMD and I think it’s fair to say that AMD’s gains would have been far more challenging had intel been competing in top form (see AMD vs nvidia in the GPU market). Both AMD and apple owe some of there advancements to TSMC’s world class fabs too and IMHO TSMC is intel’s greatest threat right now. It’s interesting to see a company few people know about at the center of it all!
Some have discussed whether intel can/should fold it’s fabs and just outsource to TSMC like everyone else, but it would lead to a global fabrication monopoly that I don’t think is good for the industry.
Yes, I think it’s good to acknowledge we don’t know these people nor necessarily have a good idea of what goes into the making of things. Good decisions, luck good or bad, and lag all have their effects. It’s a bit like politics.
There’s something to be said for spinning off a fab as management just have to focus on doing one thing not keeping a bigger corporation happy.
One reason why nobody knows much about TMSC is it’s thousands of miles away from Intel who exist in a culture of blabbermouths poaching each others staff. Simply being thousands of miles away can let you get something done. The Far-East also took on a lot of board manufacture so you have a fair amount of industry in the region involved. It’s like chess. It can be easier watching and spotting a good move than actually playing because of the distance and immediacy of the involvement. You also spotted something Chang is famous for – anticipating clients needs. He doesn’t need to follow corporate rules to aim for a design target or schedule so isn’t cognitively disabled by this process. I suspect this being close to customers in a relationship sense and more free to invent is partly responsible for TMSCs current leadership. At least, this is my theory.
HollyB,
Yeah, it’s an apt observation of politics too. One can be responsible: pay down debt and fund social programs like healthcare only to have a successor that balloons the deficit and hands out corporate tax cuts like candy. It sucks to be the responsible party: sacrifices are always made on your watch only to know that your successor is going to demolish your work and spend all the money you saved to fund them
Well, western companies, intel included, have gotten way too bloated and unfortunately it has been a long time since the US has been competitive – we’ve lost so many sectors to offshore companies. Intel had advantages of scale, that was huge, but with the proliferation of ARM devices competitors were able to get to similar scales of economy and now western companies like intel are struggling to compete against more efficient competition from the east. They could make the sacrifices to become efficient again, but US companies that were top dog for so long may have forgotten how to run a competitive manufacturing business rather than taking their positions for granted.
@Alfman
Yes all true and the UK currently doesn’t have a lot to brag about either.