News Intel to announce a 20% workforce cut this week: Report

[Admin]

Intel is reportedly planning to cut over 20,000 jobs as part of new CEO Lip-Bu Tan's push to streamline management, reduce costs, and refocus the company amid ongoing business struggles.

Intel to announce a 20% workforce cut this week: Report : Read more

 

[hotaru251]

by 2030 intel gonna be a small tech company at this rate :|

also shouldnt be able to get gov funds (ala chips act) and then proceed to gut your corpo v_v

 

[Giroro]

Slash and burn.
Rip and tear.
Strip the company, and sell the parts for scrap.
Intel's board is all-in on TSMC; they're just making it official.

 

[thestryker]

Given how damaging the 15000 headcount cut seemed to be I have a hard time believing this won't be even worse. It's quite believable that there are still problems with business structure and management, but not on this kind of scale.

As an example of the type of losses that happened Ian Cutress posted this about the VLSI conference happening:

To be presented at VLSI_2025:

A 128Gb/s 0.67pJ/b PAM-4 Transmitter in 18A with RibbonFET and PowerVia

Authors from➡️
Intel (as expected)❗
Alphawave Semi (IP company, ok)🤯
NVIDIA (what?)😱
Apple (wha-what?)

...

Update: One of the authors did get in touch. Said that a good chunk of the other authors are all ex-Intel, and said it was unlikely their new employers were involved. Out of the 13 original authors, only 6 remain at Intel.

 

[ezst036]

Someone loses their job.

Or everyone loses their job.

 

[bit_user]

before that the company is in its quite period and will not comment on material things.

Last sentence: should be "quiet period".

Also, my condolences to those affected by these cuts.

 

[jkflipflop98]

Ya know, it's always sad to see someone lose their job. But there really is too many layers of management at Intel. The company has managers that manage managers that manage managers that manage managers. Completely unnecessary.

 

[Jimbojan]

Intel does not need to lay off 20K people, its annual sale per employee is over $450K. How many companies can claim that?

 

[redgarl]

Without any breakthrough in any of their main field of business, Intel will not turn the ship around.

If they cannot become leader in any fields, it will be the end.

 

[dalek1234]

Without any breakthrough in any of their main field of business, Intel will not turn the ship around.

If they cannot become leader in any fields, it will be the end.

I think that Intel is now past the point of no-return. Thy are just rearranging the deck chairs on sinking Titanic.

 

[-Fran-]

But hey, I've been reading here from several people Intel is not doing bad and still way ahead! You know. All those OEM contracts and strong dominance?

Anyway, sarcasm aside and in agreement with @bit_user, all the best to the people impacted. Specially engineers as Intel has one of the best in the Industry for sure.

Regards.

 

[Elusive Ruse]

Damn that’s rough, especially in today’s economic climate.

 

[bit_user]

Ya know, it's always sad to see someone lose their job. But there really is too many layers of management at Intel. The company has managers that manage managers that manage managers that manage managers. Completely unnecessary.

As the article pointed out, the layoffs are greater than the total number of managers in the company.

Also, as a contrast, I'd point out that my boss's boss has like 20 direct reports and he definitely can't keep on top of what they're all doing. I think the rule of thumb is that a manager can effectively manage no more than about a dozen people.

There are other types of "managers", BTW. Project/program managers might not have any direct reports, but are responsible for overseeing large efforts with many cross-team dependencies. I think the bungled launch of Arrow Lake (and worse) is the sort of thing we can look forward to, if you don't have effective project management.

Product managers also might have no direct reports, but are responsible for integrating data from sales, support, engineering, and market trends, then distilling it down to product requirements. A bad product manager will either miss key requirements or fail to prioritize and cram in everything + the kitchen sink. I don't know exactly why Sapphire Rapids suffered an extra year of delays, but many chip bugs were reported and it has many special-function units. I have to wonder if part of the problem was just that its requirements got overloaded to the point where it became impossible to fit into the original market window.

If Intel's execution had been flawless, then these cuts wouldn't worry me so much. But, their size means either lots of products and programs will be cut, or that teams will be decimated without a plan to effectively cope. If you want to streamline operations and only cut overhead, they could not be this big.

 

[Pierce2623]

I sure hope the whole 20k comes from middle management. If so it might actually increase productivity. I’ve had a couple college friends from engineering school get jobs there and the buerocracy ran them both off, one took a small pay cut to go to AMD and one actually got a pay increase to join a startup that promptly collapsed in under 2 years….

 

[TerryLaze]

Intel does not need to lay off 20K people, its annual sale per employee is over $450K. How many companies can claim that?

Yeah but they also increased their property from 55 to 108 bil in 4 years(new FABs buildup), coming up with 50 bil in 4 years isn't easy, and then there is all the running costs of the company on top of that.

 

[jp7189]

This could be a move to pre-eliminate positions that would be redundant following a major business shift. E.g. divesting the fabs. It's better to have the layoffs under the 'old' company rather than saddling the new company with the task.

 

[bit_user]

I sure hope the whole 20k comes from middle management. If so it might actually increase productivity. I’ve had a couple college friends from engineering school get jobs there and the buerocracy ran them both off, one took a small pay cut to go to AMD and one actually got a pay increase to join a startup that promptly collapsed in under 2 years….

If you read about the history of Silicon Valley, there was time when it seemed like a lot of the founders & principals at other chip companies had previously worked at Fairchild Semiconductor. I wonder if Intel has become the equivalent sort of incubator for modern chip designers & process engineers.

 

[TheSecondPower]

Intel will have very little post-"7nm" capacity until around the end of this year. In fact this image from Intel shows 2024 and 2025 having less fab capacity overall than the preceding years. Since Intel's success always came from their fabs this is a bad place to be. Next year looks much healthier, but lots of fabs and no employees to use them could be a bad place.

AMD lost a lot of staff during its bad years and came back strong. But the product lineup clearly suffered. Desktop CPUs were completely abandoned. Some graphics card generations saw only one chip. The GPU division fell behind. Then everything changed when the finfets arrived.

 

[bit_user]

This could be a move to pre-eliminate positions that would be redundant following a major business shift. E.g. divesting the fabs. It's better to have the layoffs under the 'old' company rather than saddling the new company with the task.

Yeah, could be happening in order to make them a more attractive acquisition target.

 

[bit_user]

Intel will have very little post-"7nm" capacity until around the end of this year. In fact this image from Intel shows 2024 and 2025 having less fab capacity overall than the preceding years.

Good point. However, their chiplet strategy should mean they don't need so much capacity on the leading-edge nodes as they used to. If they can use Intel 7 for substrate and stuff like I/O & SRAM dies, then it could well complement their overall strategy.

IMO, what that graph calls most into question are the prospects of their dGPU and server CPUs. Those both need lots of die area on a leading edge node, in order to be competitive. For desktops and laptops, the compute tiles represent only a minority of the total die area.

AMD lost a lot of staff during its bad years and came back strong. But the product lineup clearly suffered. Desktop CPUs were completely abandoned. Some graphics card generations saw only one chip. The GPU division fell behind. Then everything changed when the finfets arrived.

AMD needed everything to go in its favor, in order to claw its way back. That includes Intel faltering on their 10 nm transition.

That's not to say that Intel is anywhere close to the situation AMD had fallen to, in the Bulldozer era. But, if they do reach such a point, there might be no coming back.

 

[TerryLaze]

This could be a move to pre-eliminate positions that would be redundant following a major business shift. E.g. divesting the fabs. It's better to have the layoffs under the 'old' company rather than saddling the new company with the task.

"Divesting" the FABs happened over a year ago, they made them into a separate entity with separate bookkeeping and everything.
"internal foundry"

This is just cost cutting to counterbalance the huge costs of building FABs.
Like last year they got pretty much 7 bil out of cost cutting and almost 10 bil from tax stuff.
https://www.intc.com/news-events/pr...s-fourth-quarter-and-full-year-2024-financial

 

[TheSecondPower]

Good point. However, their chiplet strategy should mean they don't need so much capacity on the leading-edge nodes as they used to. If they can use Intel 7 for substrate and stuff like I/O & SRAM dies, then it could well complement their overall strategy.

IMO, what that graph calls most into question are the prospects of their dGPU and server CPUs. Those both need lots of die area on a leading edge node, in order to be competitive. For desktops and laptops, the compute tiles represent only a minority of the total die area.

I think Intel's plans were for the server CPUs to use nearly all of the Intel 4/3 capacity, so Meteor Lake had to use very little. The Granite Rapids and Sierra Forest dies are a lot bigger and do sit next to Intel 7 I/O dies. Also I really suspect Arrow Lake-S (on desktop) and its need for high clock speed would've made more sense on Intel 3 than on TSMC N3B, so the decision to use TSMC either means there wasn't enough staff to migrate the CPU cores to Intel 3 for desktop and N3B for mobile, or there wasn't enough Intel 3 capacity.

 

[Zerk2012]

"to purge excessive layers of management, eliminate redundant positions"

Our company did the same almost 3 years ago. Revenue stayed the same but cut almost 30M off the payroll.
Most of the jobs were middle management and assistants. Our company is nowhere near as big as Intel of course.

Hate it for the people loosing their jobs.

 

[Roland Of Gilead]

Also, my condolences to those affected by these cuts.

Anyway, sarcasm aside and in agreement with @bit_user, all the best to the people impacted. Specially engineers as Intel has one of the best in the Industry for sure.

Over here in Ireland, and specifically the Leixlip plant, with Fab 34 (newest) and Fab 24, the word on the street is 5,000 employees will be cut. This is a large percentage of the total staff (I understand this to be approx 17,000), this will have a huge effect not only on those cut, but the wider support network too. As Intel makes cuts, other support companies will also be cutting staff. There are a lot of very worried folk in these parts.

Edit: Correction - I must have gotten my numbers mixed. There is approx 5000 staff in Ireland. So obviously the cuts would not amount to 5k. It will be far less. But still, this will have an impact on local economy and the wider support network.

 

[DS426]

This is wild because it begs a question: why wasn't excess management a bigger portion of the original 15K job cuts? It means both each job cut would net back a bigger return since wages are higher for managers, along with not losing as many talented and humble engineers.

My thoughts and prayers to those who have been cut or will be cut, both rounds.