Thread regarding Intel Corp. layoffs

OP... "All those who still can't see what is coming for INTC, can stick it. " is this the new tag line for long winded rants? I have seen something similar on a few long posts. Basically saying if you don't agree with me you are wrong. Nothing you have said is new so don't think that you are a genius. You left out the possibility that Intel is actually too far gone already and the end result will be to break it up and sell the pieces. Tan might just be here to clean up some of the mess so the company can be sold. Firing 35k+ people will make it far more attractive to potential buyers. Read Tan's contract, it includes provisions for "change of control" and "new ownership". Those terms aren't in previous CEO agreements. Maybe you can't see what's coming for Intel and you can stick it...

Read on the world wide web that the Board is pulling back from discussions to spin off Foundry, which makes no sense to me but apparently they are instead choosing to stumble ahead with IDM, but you know, do it better or something.

The rumors suggest that the Board thinks IDM is ok, but that they just need a more lean organization.

If this is true then it would appear that they do not accept that the company can no longer fund technology development sufficient to remain on leading node.

Maybe they don't even accept that x86 is now in long term decline, overtaken by ARM and soon to be attacked by RISC-V.

If any of this is true then the Board is delulu and Lip-Bu may make the org more lean but none of the strategic issues will be resolved. I was hopeful that they were ready to call it quits on IDM, as everyone not carrying a blue badge knows it is a failed strategy.

It will simply take too long (maybe a decade) to build enough external customer demand to keep the Foundry capital cycle going, and it can't be spun off (for much of a valuation) while posting huge losses.

Not trying to be negative or a doom-poster, but see if these rumors play out, because that means in a year or so the company gets picked apart.

I always though that IDM 2.0 was just a ruse on the manufacturing side, to rebuild Foundry capability so that it could be spun off. That may still be the plan, but of course the Board will never, ever say that quiet part out loud.

Tech company strategy is never stated. It always has to be inferred.

Tan has a VC background he's going to cut things that are not efficient and projects that don't make money!

A bit amusing to read so many posts about 'someone else' being terminated, largely written by Fab workers who are the very most likely to be let go or turned into contract workers.

I'm sure some managers and even some ELT will be let go, as part of an effort to streamline the organization, and some workers of all levels will be let go from product groups which are shut down or sold.

But that is not where the real fat resides. The real excess headcount is in the fabs, which are still to this day organized the same as they were when 'Max Output: leave no wafer behind' was the strategy.

This is not how any other legitimate foundry is organized. To be cost competitive with TSMC, Samsung, or even GF, those who actually need to be in the fab are converted to contract workers (far fewer benefits but decent pay), and EVERYONE else is no longer a position specific to a fab.

Virtually all those other workers are either offshored to COE-like processing centers or if the role is even maintained in the US, then it is centralized to Arizona or something.

TD fabs might keep more people on site but even there they will lose headcount.

No one will invest in the current money-losing, grossly inefficient IFS, and that makes restructuring it the highest priority, because spinning out IFS is the most important strategy. Product groups can no longer fund the capital requirements.

The current wafer cost is so high that Intel would lose money on any external customers it gets, until that cost is made competitive to TSMC. So maybe it is a good thing that bringing major customers onboard has gone slowly?

No @cp+1jp67g9b3, Twitter lost 70%+ value after Elon.

@aw Hey!! I resemble that remark!

Some good points. I'd just say annual rank-yank will ki-l One Intel, and make us too myopic. Much rather cut deep than drag it out, or use disingenuous process. Microsoft was notorious for their rank-yank, and they did a big u-turn after dropping it...

Also would caution going full Elon on the cuts, that can be value destroying. Look at Twitter.

For IFS, raising customer capital >> buyouts. The geopolitical environment is favorable. Also Intel would have to factor the AMD cross- licensing poison pill, and that'd be a lot of lawyering distraction from the engineering. Good thing I think the stock is already 'underpromised' to investors. Now we just need to over deliver

All of you know that you are just employees at Intel, right? Each and everyone of you are d-mb enough that you are not in any form or shape capable of making any decision for the company or its future? Now let that sink in.

He is coming to chop up and sell the parts

@a8 The comment wasn’t that an Intel DC GPU isn’t ’already great’. Yes, Intel should’ve stayed committed to expanding beyond CPU. Fact is, it didn’t, and someone else got a two decade head start.

Now they have the better software, op model, engineers, brand recognition, and relationships in the industry. Intel doesn’t get back in the data center until Quantum.

Didn’t the last CEO do toxic optimism enough for you?

@aj+1 CPU is not going anywhere, unless sold to Broadcom or Qualcomm.

It is still very profitable as a legacy technology, even if it will likely never again provide big growth. ARM is going to take more and more share, and the company ought at some point get into RISC-V, as that architecture is also starting to produce server products.

AI is the big wave and the company needs to improve (by a lot) it's ability to understand where that market is headed and what product is best suited to be competitive.

The current AI efforts are still subpar, and AI consists of several TAM, so there is plenty of room for Intel (and AMD) to find profitable niches. But not if they don't try.

Ki-ling CPU is nonsense. CPU is 95% of revenue and much more than 100% of profit, and it subsidizes everything else. There is nothing to show for GPU after decades of investment and tens of billions of dollars down the drain.

I agree that headcount and products will be streamlined but he is not some wild chainsaw dude.

He has a Finance background, although he did have an engineering degree (but not directly related to semiconductors).

Better to think of him akin to having Bain capital as CEO.

I think he will develop a more clear strategic direction, then go about making the company profitable (that is where the pink slips happen) once he has what he and the board consider to be the best products for current and future markets.

The Board and ELT have been difunctional because of the lack of shared commitment demonstrated in so many poorly chosen products, and some good ones, which never lasted long enough to find a market even if they were good. I expect him to clear that up, then cancel or sell off the misfit products and groups.

Every tech company is challenged to develop products for markets and niches that often don't currently exist or are in some infant stage of development. AI has shown how important it is to be right and on time, because new technologies are happening much faster than they used to.

AI is going to cause that to accelerate even more, making product selection even more winner-takes-all than it already has become.

This is less of an issue on the manufacturing side, although they have their own challenges in what has become the most difficult engineering effort on the planet.

If Lip can restructure Intel Board and upper management to be lean enough to make the right decisions then that by itself would be a huge win. That means slicing up the company vertically and horizontally. There are simply too many people involved in strategy for it to ever be successful. Koduri gave an interview which touched on this, where he referred to these leeches as powerpoint snakes.

The conglomerate needs to be simplified, shedding many groups which no longer are suited to the core strategies in CPU and AI.

Manufacturing needs a similar downsizing, getting rid of many layers of management and stopping all the redundant functions across fabs. By redundant I mean any person who does not have to be physically on site, because that means that what they do can be done by some person who works for more than 1 fab, and can often be based on a low Geo country.

All this was going to happen if the Broadcom and TSMC rumors panned out, because it is the only way Intel can survive as a going entity.

The rent is too damn high!

GFx engineering team in Intel is 3rd tier. They make up for it with inferior management :)

That group will need a good scrubbing, flushing, and reinvesting if it is going to compete with nVidia, the hyperscalers, AMD, etc.

15k reductions minimum, and that is before the split.

Get rid-off CPU (Atom, Xeon) and 50% reduction of IFS/TD/TMG, PEs/Senior PEs. They can't find jobs outside Intel. Just hanging around via nepotism and brown nosing and collecting $500K+ compensation

Some observations:
+ Engineering orientation
+ Cadence seems to get positive reviews for flexibility on Glassdoor, and has some WFH. Hope he won't go Elon on us.

As long as he eradicates HR first, I’m good even if I get the boot.

@a7+1 I think OP was really referring to any type of AI processing vs CPU.

It seems CPU has been potentially permanently deprioritized in the datacenter, and ARC has shown there is plenty of room on the consumer side to compete in various segments.

No company is instantly awesome. The winners get there by being persistent.

Also, this site is infested with NVDA and TSMC fanboys, who aren't really here for a meaningful discussion. Fair enough, those companies have been the winners, but that can change.

Add to your list that the vast severances which are coming will not include the generous packages which were previously provided.

A. There is no money to pay for that.
B. Such packages are not the industry norm. Lately terminations from other tech companies got essentially nothing.

With this in mind, everyone should be dusting off the resume and starting to search for their next role.

Floundry has previously escaped the deep cuts but that will no longer be true. They have been preparing to eliminate a large number of positions which are redundant across fabs, similar to what happened when they ramped up ROC.

Some product groups will be sold. Starting with Networking and any other groups which really only existed to help Intel sell a while datacenter package. That stopped making sense when Intel lost dominance in datacenter and that will be GPU based for a long time to come.

I'd agree about the need to make GPU the highest priority and the comments to the effect of 'it's not already great' typify how dysfunctional the Intel mindset has become.

This is exactly why the company keeps starting and stopping products, which other companies relentlessly pursue because those companies understand what it takes to match products with future markets.

Anyone with this mindset is incompetent and needs to be terminated for cause.

Inference requires dGPU OR integrated acceleration (ahem, NPU) because gains model in efficiency and hardware are happening faster than increases in compute requirements for inference workloads.

Turning Intel back into a “growth company” was a choice of either IFS or data center GPU. Both had low odds of success. Failing at another data center GPU wouldn’t have been fatal to the company at least. It was late to start then. Way too late now.

@a3+1 Probably right about inference but that is also using GPU.

The company needs an offramp from CPU, and is making progress on GPU.

They mostly seem to need clarity of purpose in the executive ranks and I think Lip will help in that regard.

Time to terminate all the backstabbing executives, or risk more floundery.

Lol not competitive in gpu

@a1+ I'm OP and said nothing about the existing products being amazing, but you seem sensitive to the reality that Intel can get better once the whole organization is working towards that goal.

GPU is likely the best path for future growth, it the company can achieve improvement.

Why tf do you think Intel should go chase Nvidia in data center GPUs? Quantum will be here before Intel is able to overcome Nvidia’s CUDA moat. The datacenter battle is won for the next decade.

The play is to optimize for inference. Even Nvidia says it.

There is a whole page of comments on this site which are not going to age well.

Company has a lot of work to do, but Lip should be able to get the C-suite aligned (or terminated, either way is fine) to execute strategy.

Getting the company cut down to size and eliminating excess decision making layers will help a lot.

Firing the executive traitors will help get everyone pointed in the right direction.

OP is the delusional engineer who thinks Intel had amazing Gfx products and will make some even more amazong Gfx products in the future. Get real you id--t.