I can't think of a single person from the past who qualified either. If you're thinking of Grove, I don't agree.
Gelsinger “retires”
- Thread starter blueone
- Start date
Sometimes engineers, even the most senior ones, make misguided recommendations. In the end the CEO has to be the ultimate decision-maker and take responsibility for success or failure.
It's only temporary.
tooLongInEDA
Moderator
At this point, it's hard to believe anyone actually knows.
Pat at least had a consistent strategy (even if some of us weren't totally convinced by it).
Has having co-CEOs ever worked ?
Doubtless a lot more to come out. But starting to feel like the headless chickens of the Intel board have taken over (I hesitate to use the words "taken control" since they're as much at fault as anyone for the last decade of failure and chaos).
delong.height
Active member
Well, i think if the engineer team should be able to estimate how that technological breakthrough would gain compared to last-gen, And how the cost would be. I think a company as big as Intel can definitely do a certain level of estimates, right?
delong.height
Active member
Tim Cook is not a engineer, but he can still right the ship. M series chips is definitely not under Jobs watch, neither does vision pro and airpods. Nevertheless they can still achieve many of these chip companies can't.
Two examples from Intel's past prove otherwise. Itanium and Optane. Optane, or 3D XPoint as it was called internally, might be the best example of how engineers can go astray. Unless you have basic expertise in a field, like fab process or computer engineering, it is difficult to see past a passionate but political technical leader.
Chip design is far easier to manage and judge than chip fabrication. I would argue chip fabrication is one of the most difficult manufacturing problems on earth. Chip design has results even non-experts can be trained to understand. Apple's M-series CPUs produce product results which are understandable by non-experts.
Anyone in the manufacturing line below the point of the pyramid is under massive pressures constantly.
Maybe those cranking out the designs need to have a word with the end fabricator
MKWVentures
Moderator
Pat tried Fab Strong. It didnt work. Now someone has to pick up the pieces of the disaster of billions in wasted spending and selling off assets to pay for foundries people do not want. I said that Stacy would watch for a while.... but it wasnt getting better and the wildly optimistic/unrealistic comments needed to stop.
The Intel Product group is successful and worth 150B alone. The IFS group is worth some negative number.
bashee_wang
New member
Agreed. Considering the pay difference, why cost so much with CEO if to trust engineers?
coldsolder215
Active member
I don't think it's fair to pin Optane on the engineers, arguably it was doomed off the bat when they kept it closed-source to justify even higher premiums for their ecosystem instead making it accessible for outside developers to leverage its strengths. Rather than proper (slow) R&D with managed expectations, they started developing products assuming the technology would eventually hit certain specs. When that didn't happen fast enough it devolved into a classic pump and dump. They announced the "breakthrough" and made the Optane guy a senior fellow well before they had viable yields. All top-down decisions culminating in the engineers holding the bag.
Could they appoint the CEO of Mobileye as the new CEO? It would send a strong message to the public about their focus on AI. At the same time, they could negotiate a deal with TSMC to address its fab issues. If they plan to spin-out the fab, now might be the time to do it.
So this announcement reads like it’s quite focused on Product. Perhaps with disasters like the Arrow Lake launch and their GPU business in the gutter still that was the last straw?
That being said, Zinsner leading the company doesn’t necessarily mean they’re going to go Fab-lite as he actually led a successful manufacturing turnaround at Micron. I do think he’ll stop trying to make the next “Silicon <random cow pasture in Ohio/Germany/Israel>” and focus on having actually profitable operations out of Arizona and Ireland.
Crazy news but not that unexpected… tough times to be working at Intel…
That being said, Zinsner leading the company doesn’t necessarily mean they’re going to go Fab-lite as he actually led a successful manufacturing turnaround at Micron. I do think he’ll stop trying to make the next “Silicon <random cow pasture in Ohio/Germany/Israel>” and focus on having actually profitable operations out of Arizona and Ireland.
Crazy news but not that unexpected… tough times to be working at Intel…
siliconbruh999
Well-known member
I want some whistleblower to blow the whistle.
Have they considering fired the Board of Directors after firing 3 CEOs?
Have they considering fired the Board of Directors after firing 3 CEOs?
lispwright
New member
Is that really set in stone before Intel takes money from the government?
Just woke up to this news, and am wondering why now? The timing a few days during a holiday period after the public announcement about Granulate is strongly suggestive.
As far as we can tell it was a complete failure of PG's investment of $650 million the company could really use now for new fabs. It's something easy to understand at a high level, something a board faction could use as a last straw argument. Even if Intel has long proven they can't acquire companies or post abandoning memory have any big successes outside of x86 CPUs, which should prompted the board to say no to it in 2022.
It is obvious PG was piling up too many failures, and someone may have also noted the unnecessary ecosystem alienation. AI chips were in his area of expertise of design, although one wonders if anyone with any power in Intel understands the ecosystem challenge of CUDA. It's certainly impossible for AMD to understand well enough to really fix.
The biggest, and one any board should be able to understand, is that PG failed to manage financial expectations, his team didn't realize or accept the very public boom in AI hardware spending hurting their datacenter revenue by simple displacement. I think that boom has been going on long enough before the big financial forecast miss which created a wide and perhaps correct perception of doom.
The temporary setup may not be terrible, the last CFO to CEO elevation wasn't all bad. Bob Swan realized the fab part of the company was failing hard with no success in sight, and told everyone in the company they were dispensable by booking a lot of TSMC capacity, perhaps a goad that worked?
If the company continues PG's strategy, we could assume Intel's technology problems are as much under control as possible today, e.g. without a major attempt to fix or adjust the dysfunctional culture which would require firing a great many managers up to the highest levels which few people are mean enough to do. So going forward the best possible financial execution of fabs gamble is what's called for.
lispwright
New member
That happened! I learned on this site the only remaining board member with industry experience resigned in disgust.
It's then a high level observation to note the board doesn't have what it takes. And I've long wondered why they hired a design guy when fab execution at all levels was Intel's existential problem, assuming the board was dissatisfied with Swan moving to TSMC and hired PG with a plan, I assume supplied in part by him, to recommit to fabs.
UMC post exiting the leading edge. But Taiwanese companies have strong chairmans (unlike most modern American companies) to act as something of a tie breaker. Technology development and manufacturing is under SC Chien and customer engagement/design enablement and strategy are under Jason Wang. Personally I like the org structure and I wonder why it isn't more common.
Agreed. The BOD needs to pick and commit.
There is no evidence of that. Every product released over the past 3.5 years was defined and spent most of its development during BK and BSs tenure. Pat didn't choose to outsource chipsets to Samsung. Pat didn't choose to outsource a whole generation of Intel products and give a slow volume ramp to Intel 4/3. Nor was it Pat's choice for Intel to have only build two development and one HVM fab over the entire tenure of the prior 3 CEOs over the past 13 years.
Interesting that foundry losses were planned to evaporate right after Intel ramped products that that started development under Pat and primarily use Intel foundry. Crazy how that works.
Lower margin than their fabless peers with shrinking market share and revenues. They also are selling 2/3 the unit volumes they promised to investors. Finally Intel products has literally failed to enter every single market they have ever tried to enter other than server CPUs over the past 40 years. I don't know how anyone can look at Intel products and say they are on anything but a downward slope. Their golden years were off the back of a pair of crutches known as TMG and margin stacking. Meanwhile Intel foundry is by all expectations going to pass TSMC next year (technologically) and gets multiple new external foundry customers every quarter for the past year or two with a glide path to profitablity in 2027 even off the assumption of immaterial foundry revenue. GF needed over a decade and to end their leading edge development to break even, and they needed 3 years to secure a non AMD non chartered customer (and even the. it was just second source production from STM who was licencing design compatible IBM tech). Samsung needed 6 years to land QCOM and by all indications is still not profitable and has less revenue than Intel foundry even after Intel moved a large portion of their products external.
I don't agree about fairness to engineers. Optane was conceived as a memory technology, not a storage technology, but with limited write endurance. Better than flash in every spec, but to compete with or complement DRAM you need nearly unlimited endurance. The engineers tried to mitigate this with a DRAM cache, but it wasn't a sufficient fix. Also, Optane layers, called decks, are very expensive to engineer and fabricate. So except for specialized applications (like write buffers for transaction logs) it made Optane SSDs too expensive for high volume applications. Good technical leaders would have painted an accurate picture.
Agreed about the making it proprietary mistake, at least for memory. You can do that with some technologies, like system-internal networks (e.g. NVLink), but for memory products that did seem unworkable. But Intel at the time had about 90% or more of the CPU market, and I suppose that colored management's view on proprietary practicality.
That senior fellow you're talking about, Al Fazio, should have been more forthcoming about the technical challenges. Intel at the time was wrestling with DRAM getting an increasing share of wallet on client and server systems costs. It is easy to see the business story for Optane, but the engineers were selling a strategy that was widely questioned by other Intel engineers. What I never heard, until they de-committed, is what Micron thought of Intel's strategy.
Overly optimistic behavior at the C level does not always work. Especially for a company with old school culture. An Intel pivot is coming, absolutely.
I just don't see why companies fire CEO's without replacements unless one of the CO CEOs will take it. In that case I do not think this will end well.
Who are possible replacements from the semiconductor industry?
I still say that Hock Tan should acquire Intel and pivot the hell out of the company. Hock could run a foundry, absolutely.
The GSA awards are on Thursday. Hock Tan will receive the Morris Chang Excellence Award. I will visit the Intel table to see what the temperature is for sure.
I just don't see why companies fire CEO's without replacements unless one of the CO CEOs will take it. In that case I do not think this will end well.
Who are possible replacements from the semiconductor industry?
I still say that Hock Tan should acquire Intel and pivot the hell out of the company. Hock could run a foundry, absolutely.
The GSA awards are on Thursday. Hock Tan will receive the Morris Chang Excellence Award. I will visit the Intel table to see what the temperature is for sure.