Intel CEO Blames Predecessors For Manufacturing Woes (axios.com) 57
When it comes to Intel's recent manufacturing problems, Intel CEO Pat Gelsinger places the blame squarely on his predecessors -- many of whom he notes were not engineers deeply steeped in chip technology, as he is. Axios: Gelsinger has announced a broad plan to reinvigorate Intel by doubling down on manufacturing. However, the strategy depends on the venerable semiconductor giant recovering from recent stumbles. Gelsinger told me that the company had grown so successful that leaders wanted to move the strategy away from what had made Intel a chip juggernaut. Especially lacking, he said was the "maniacal" focus on manufacturing that had been a hallmark since Intel's founding. Gelsinger returned to Intel as CEO earlier this year, spent three decades at the company after joining it at age 18.
Re: (Score:2)
Re: (Score:2)
Get fucked.
https://medium.com/@rikimatsum... [medium.com]
3 Envelopes (Score:5, Funny)
Re: (Score:2)
This was my first thought on seeing the headline. I am got it got frist post.
Re: (Score:2)
Mine, too. Came looking for it, and was not disappointed.
Intel CEO Blames Predecessors: That seems correct. (Score:3)
Three of the MANY examples [slashdot.org], in the years 2006, 2014, and 2019, are given in a comment I posted last year.
Re: (Score:2)
He opened the first envelope (Score:5, Funny)
It seems he's heard the story about the new CEO with the three envelopes [dailyjokes.co] too.
He's not wrong. (Score:5, Interesting)
Joke about the fact that he's pretty much following the trope of every CEO ever, but Intel did make a lot of mistakes. They wanted to diversify the company into becoming a software and other stuff company instead of focusing on improving their processor. I am quite certain I commented that this was going to be a huge mistake for them right about the time they bought McAfee. AMD was doubling down, ARM was becoming more and more entrenched in the mobile space and Intel had dominated for so long that they left the processor divisions on auto pilot while they stumbled on trying to reduce die size.
Re:He's not wrong. (Score:5, Insightful)
Re: (Score:2)
This is why the second envelope exists... Reorganize.
Sadly, that only leaves him one envelope left.
Re: (Score:3)
Blame game. (Score:2)
Sometimes even engineers can be part of the problem. [hbr.org]
Re:Blame game. (Score:4, Interesting)
That is an interesting article, I wonder how long it will take before Apple takes its place in that pantheon.
Intel have been here before - when AMD first moved to energy efficient 64-bit processors and left Intel's Pentium 4 and Itanium in the dust. Intel managed to regain the lead with families of new processors while simultaneously engaging in anti-trust violations to edge AMD out, AMD barely survived.
Re: (Score:2)
"I wonder how long it will take before Apple takes its place in that pantheon."
You left out the word again. Apple was in serious trouble before the second coming of Steve Jobs. Then they got in trouble again when IBM couldn't get the G5 in a laptop, and Motorola wasn't interested in upgrading the G4 processor. Hence the jump to Intel.
Obsession with ROI has (Score:5, Insightful)
screwed up a lot of companies. Certain industries require a relatively long-term outlook, but modern finance practices dictate "get it now and F the future". IBM and GE both clobbered themselves this way.
It does "work" in the short term, but milks the company dry such that they have no ammunition for the future. It can damage your innovation portfolio and reputation. For example, IBM damaged their customer service reputation by being a persnickety cost cutter. It's hard to fix a bad reputation, as the Detroit automakers can tell you.
Ford busted their butt to make sedans with quality comparable to the Japanese brands. However, surveys showed that too few shoppers believed that. They'd probably have to beat the Japanese brands for about a decade to get their reputation back. Breaking even was not enough. They realized it was a lost cause and decided to get out of the sedan business, except for the Mustang where shoppers value sizzle more than quality.
Re:Obsession with ROI has (Score:5, Informative)
Clarification:
RE: "Ford busted their butt to make sedans with quality comparable to the Japanese brands."
This was after poor quality in the 70's, 80's, and 90's.
Re: (Score:2)
Re: (Score:2)
And after that. When renting cars, I have had the displeasure of having to settle for Fords when there were no other options. The lack of quality after the year 2000 was noticeable compared to even GM. With auto rental companies, they generally get the cheapest models with the fewest options so my experience might be skewed.
My experience has been more Dodge than Ford. It wasn't until about 07 that Ford decided they were going to revamp their processes.
I remember hearing a story about the new president of Ford "buying" his mother a car. He supposedly sent her to the Lincoln dealer, told her to pick out what she wanted and he'd get it for her. She called him immediately after going to the dealer in tears because she was overwhelmed with the options available. Sure enough, the car she was looking at had so many options that t
Re: (Score:2)
I love that you guys are arguing over which American car brand produces the worst cars. It's like arguing who should be at the back of the short bus. And the answer is clear - all the British car companies have you thoroughly licked in terms of producing crappy quality cars. Even Tesla can't manage a panel gap like a British Layland.
Re: Obsession with ROI has (Score:2)
Its spelt Leyland and it went out of business years ago. JLR however build some of the best cars on the market right now, far better than any crap Ford can vomit up.
Re: (Score:2)
Well, Chrysler had the K-car in the 70s. What a piece of crap that was. They figured at the time that the Japanese secret was in economy cars, so they indeed built a cheaper car. It was cheap in price and cheap in quality. They hadn't figured out yet that the Japanese were making inexpensive but reliable cars.
Re: (Score:2)
Unfortunately for Chrysler after Iacocca saved the company, he handed the rains over the Bob Eaton who then "merged" with Daimler at a time when Daimler was cheapening Mercedes.
Re: (Score:2)
Well, ours was in the shop all the time. Essentially it had a bad carburator, but they did not want to replace it so they did lots of tweaking. It was all under warranty so we didn't pay for it, it was just a hassle and they should have just done the expensive repair once. When warranty expired we dumped it fast. Amazingly, the parents stuck with the same dealer that did the bad service, instead of dumping them too.
I don't know if they sold that many K-cars, not enough to save the company I wouldn't hav
Re: (Score:2)
Re:Obsession with ROI has (Score:5, Insightful)
ROI and investors really screw up companies. Investors don't even know how things work on the inside, when they start calling the shots they can really be short sighted. Either way shortsightedness from seeking after ROI can really screw up companies in the long term.
Re: (Score:2)
In some cases it makes financial sense for an investor to squeeze the worth and wealth out now to make a quick buck, and then dump it if it doesn't survive the squeezing. Their wealth algebra may factor in the fact that a certain percent of cows will get really sick if you robo-milk them.
Re:Obsession with ROI has (Score:5, Interesting)
The tale I hear repeated over and over from various friends working at Intel goes like this:
1. Intel realizes they need to diversify beyond CPUs.
2. Intel spends big $$$ to go into a new semiconductor area (Altera, cellular modems, routers, wireless widget, etc).
3. Engineers get the thing working, maybe not blowing their competition out of the water, but still saleable, likely the envy of any business person outside of Intel.
4. Bean counter looks at the 5 year forecast, compares it to ROI of the very profitable processor side. Sighs.
5. Intel exits the business lays off most of those involved.
So folks are justifiably hesitant to work on anything not in service of the CPU business, and the bold thinkers are left cranky or out of work. Lather, rinse, repeat.
They also keep trying to get into the external fab business, while even internally anything other than CPU's goes outside to TSMC and the like. Design rules are impenetrable, and nobody in the fab will give you the time of the day if it is not in service of the CPU business.
Most of his predecessors WERE engineers (Score:4, Insightful)
Only Paul Otellini and Bob Swan were not engineers, and the latter wasn't only in the job a short time.
Remember it was under Craig Barrett's watch that they made a) the Pentium 4, a desktop processor so bad they were lapped by their mobile processors that were still based on the Pentium III, and b) the Itanium, an 64-bit processor that was so bad they were lapped by the Pentium 4.
Re: (Score:3)
I see a lot of similarities to the bad old days of Pentium 4. Some 11th gen chips are slower than their 10th gen counterparts, AMD is mopping the floor with Intel in terms of performance, and Apple is using its own silicon again.
The more things change, the more they stay the same.
Re: (Score:2)
they made a) the Pentium 4, a desktop processor so bad they were lapped by their mobile processors that were still based on the Pentium III
I agree the Pentium 4 was bad (the correct answer is to agree), but Pentium 4 was introduced over 20 years ago.
Re: (Score:3)
"Pentium 4 was introduced over 20 years ago"
Not only was it still Intel, with an engineer CEO, Pat was there at the time.
Re: (Score:3)
Pentium 4 wasn't so much about mismanagement as it was engineers just betting on the wrong horse. R&D is hard for many reasons, one of them being you can't see the future and you don't automatically know what path is the right. There is a ounce of luck involved. Intel thought the way software was going would benefit from high clocks and lower IPC. Turns out they were wrong and AMD's more traditional design performed better in most ways that mattered. The fact it was also 64bit looked cool on the packagi
Re: (Score:2)
Even if you want to give them the benefit of the doubt for Willamette/Northwood, they still doubled down for Prescott AND tried to do a process shift at the same time. They moved DPG into the same building as LTD Design and put both on Prescott and it sill took eight full-layer steppings to get to saleable product. It wasn't until Cedarmill that the actually got the performance that Willamette promised, then they killed Tejas at A0 convergence. They didn't tape-in Tejas.
Re: (Score:2)
And if they had dropped it at Willamette they wouldn't have had a desktop product for years. Core couldn't have come that much faster than it had.
Re: (Score:2)
"Core" is just Hebrew for "Pentium III". IDC was told to target only mobile platforms and they still wound up with a desktop capable processor. They could have been given a deliberate desktop target at any time.
Re: (Score:2)
Core is descended from that lineage but there are many changes a long the way. You can't simply call the first Core Solo chips 65nm refreshes of Pentium III.
Blaming predecessors (Score:2)
Nikita Khrushchev assumes total control of the Soviet Union after Stalin's death.
The first time he sits behind the dead dictator's desk, he finds three envelopes, addressed to "My Successor, open in case of problems" numbered 1, 2, and 3.
Things are humming along but afyer a while, the economy tanks, there are revolts in the west, so he opens the first one, and it says, "The first time you get in trouble, blame your predecessor." So Khrushchev gives his "secret speech" to the politburo condemning Stalin and
The product isn't the focus, the shareholders are (Score:5, Insightful)
It's kind of amazing that we are in a point in this dystopian nightmare where a CEO has to announce that the problem with the company is that they didn't focus on the product they make. The company would seem to exist to make chips but, in 2021, it obviously doesn't - it exists to be a money factory for shareholders who know everything about the company's numbers and absolutely nothing about the actual product.
I don't mean to under value the business skills it takes to run a huge company, I couldn't do it, but it is amazing how many executives are filling roles which they have absolutely no background qualifications in whatsoever. The biggest place you I see this now is in CIOs who do not understand the first thing about technology. Astonishing levels of ignorance. They know how to keep the budget numbers looking good, right up until the inevitable disaster that results from tech debt.
Re: (Score:3)
All companies exist to be money factories.
The problem is, the grass is always greener. GE existed to make money by building... things. Then they started a financial services division, and profits are so stupid in financial services that it became the tail that wagged the dog.
Intel existed to make money by making chips. Then they saw the stupid profits in making software and....
The temptation exists all through a modern economy. Services generate big profits because you don't have to worry about all the logi
All recent INTC CEO's mistakes were ironic (Score:5, Interesting)
Snow was over his head technologically and didn't know better -board should never have picked him but he probably thought he was doing a good job right up until the end, ironically. (Had they come up with the right inducement for Gelsinger instead of Snow back then, things would be far better now)
Krzanich, screwed up manufacturing because he CAME from TMG and thought he'd usher in a new era of Intel dominance by insisting on risky design features for 10nm which he was unwilling to back off from until he was forced to. The person most at fault for Intel's process problems.
Otelini, while generally well thought of for green lighting the Left Hand Turn strategy switch to low power Core architecture will admit that he made a terrible mistake by not taking Apple up on the offer to do phone processor for them - ironically because his hardheaded sales/business background told him Intel would lose money making that small volume product for Apple.
Barrett, who was a semiconductor engineer, thought he'd ride the dot-com boom into the 2000s by making a bunch of expensive and tangential network device acquisitions that diverted Intel from core semi competence (in an attempt to be Cisco too) and gave us the crutch of Pentium 4 and the mindless obsession with frequency. Ironically (again) compromising good technological directions to pursue merely marketable features - which high frequency WAS - is something you'd expect his sales-oriented successor, to do. Instead, Otellini was the one to clean up that mess.
Andy Grove was the last great Intel CEO until (we can only hope) Gelsinger steps into that pantheon.
Re:All recent INTC CEO's mistakes were ironic (Score:4, Interesting)
I note that you give no examples of errors that Bob Swan made, as opposed to the others.
The fab processes were already screwed up when Bob Swan took over. He correctly promoted 14+ and 14++ which gave the circuit designs a chance to rev with less process churn instead of sandbagging on timing and yield.
Andy Grove promoted "constructive confrontation" but in many cases to failed to remind people of the "constructive" part. Andy was gone by the time I worked there, but there were people that he put in place that were merely assholes.
Re: (Score:2)
Otelini - his hardheaded sales/business background told him Intel would lose money making that small volume product for Apple.
He seemed utterly obsessed with not losing money, even in small amounts. That's not a great strategy for a company that depends on innovation.
Re: (Score:2)
Otelini, while generally well thought of for green lighting the Left Hand Turn strategy switch to low power Core architecture will admit that he made a terrible mistake by not taking Apple up on the offer to do phone processor for them - ironically because his hardheaded sales/business background told him Intel would lose money making that small volume product for Apple.
Well wait, though ... when has Intel even proven it has the wherewithal to make phone processors? Last I heard, they couldn't even make cellular modem chips that any phone makers really wanted.
BS like raid keys and less pci-e lanes on workstio (Score:2)
BS like raid keys and less pci-e lanes on some workstations for the lower end suks when the last gen did not do that are ripping of the end user.
Pudding (Score:2)
Um, the exception proves the rule? I was going to comment that it's the usual finger pointing BS, but no, he's right, when the bean counters get put in the charge, everything ossifies and you are on the long road to perdition. Look at Xerox, Bell Labs, IBM, Kodak.
Hit The Ground Running - Engineers (Score:4, Insightful)
There was a book I read awhile ago called Hit The Ground Running. It was a smallish book that was the writer's summation of research he did to try and figure out the nature and practices of some of the most successful CEO's in the USA. Granted it was his definition of successful, but after reading the book, it seemed he picked winners. One of the most interesting things I took away from it (and maybe I remember it wrong... but I don't think so), is the majority of the most successful CEOs started as engineers. I think engineers are good problem solvers, and look at data differently; looking to understand the biases, understanding SPC, etc. And the better engineers in terms of being CEOs were good people persons. But overall trying to make the companies function better as a means to improving long term profitability. That was my take.
Re: (Score:3)
There was a book I read awhile ago called Hit The Ground Running. It was a smallish book that was the writer's summation of research he did to try and figure out the nature and practices of some of the most successful CEO's in the USA. Granted it was his definition of successful, but after reading the book, it seemed he picked winners. One of the most interesting things I took away from it (and maybe I remember it wrong... but I don't think so), is the majority of the most successful CEOs started as engineers. I think engineers are good problem solvers, and look at data differently; looking to understand the biases, understanding SPC, etc. And the better engineers in terms of being CEOs were good people persons. But overall trying to make the companies function better as a means to improving long term profitability. That was my take.
I wonder if that's less a result of the qualities of engineers and more a result of natural career progressions.
For people who start on the MBA/management path there's a clear progression from entry level to the C-suite. It's just a question of how fast your promotions go and where you level out, which leads to a lot of emphasis on politics and luck on the org chart.
For "engineers" (including other technical types) the natural progress is more senior technical roles, or team lead roles, but there's not that
King of the hill (Score:4, Insightful)
Time and again companies who make it to the top fail before long. Monopoly positions allow rot to fester, make it easy to not be hungry, they sit on their laurels having slayed all the beasts in their path. Life is good. Risky ideas are dismissed as slaughtering the golden goose, just have to stay the course and profits will roll in. Often this works for many years.
Competitors get hungry, and see a fat goose ripe for disruption. Frustrated employees are easy poaching targets, so the innovators leave while those with no skills beyond turning the same old crank entrench.
Before long the lumbering giant finds its monopoly position in jeopardy and scrambles to become a lean mean fighting machine again. Problem is that your best talent either rotted, or left. New talent has been steeped in the slow moving risk averse ways from day one. Trimming a slow moving bureaucratic mess just makes a smaller wounded one, good talent avoids companies in the middle of upheaval and layoffs. Lower costs and a continued market share advantage buy time, but inevitably the organization can never go back to its glory days. At best you get a new normal, but history is littered with zombie companies that just shamble on, often as little more than a brand name as part of a conglomerate that milks the long tail for profits to leverage into the next acquisition.
Old Joke (Score:2)
A new CEO takes up his position and finds in his drawer three numbered letters and a note. The note reads:
"Dear successor,
Congratulations on your position. Enclosed you find three numbered letters that will help you in dire times, when the economy isn't working out, when the sales numbers are not up to expectations and when the board bothers you. Use them in that order and that order only. Best wishes"
Well, he starts in his position, but the whole thing doesn't exactly pick up steam. And quickly the board d
Taking his queue from Washington politics, (Score:2)