Thread regarding Intel Corp. layoffs

Why Intel failed - a view of a L0 employee of 24 years

I have been a L0 employee for 24 years before I left in last October's layoffs. I have observed the managers and management over the years at times wondering if decisions being made are right or, if the right people in the right positions but never lost trust in management. Even after the badly executed ACT in 2016, I still trusted the people at the top and assumed they are better informed than me. But gradually it was becoming clear to me that decisions being taken at the top are ill advised and the Executive management is incompetent to take sound decisions. I wanted to list key events and decisions that made Intel slide downwards incrementally.

2003 - AMD64: It was in the works for a while and Opteron CPUs were released in 2003. Intel (I mean the management) were obdurate and stuck to the line "PCs don't need 64-bit computing" & "Itanium is superior". Itanium was around for a good few years by then to be able to judge if the architecture will ever gain momentum. But yet, PSO determinedly said "Intel never lost an architecture battle". This mistake opened the doors of data center market to AMD.

2006 - SET: Bain or someone advised PSO to separate people management & functional ownership. This decision looked innocuous at that time but has had the most devastating effect on Intel. Career managers got more prominence than technocrats and lead to weakening of technical competence in the decision makers. Machinations by the managers lead to even key technical roles being filled by those with good soft skills but questionable technical achievements. A habit of ignoring technical trends in favor of short term goals took root after this. E.g. declining XBox business, foregoing iPhone business. Where strategic decisions were made, they were found wanting - e.g. Selling ARM business on the premise that Atom can fill the gap, Buying McAfee or betting on half baked Nervana. All were result of the decision makers being advised by some very incompetent staff sorrounding them. The benefit headcount reduction was rolled back soon as PSO announced a revenue target of $70billion and the newly minted empire builders used the chance to take headcount to 100k+ and keep it there. Deterioration of Intel's culture started and has continued.

2012: AB becoming chairman of the board: The pretty much made board oversight useless for vetting strategy and vision. Bean counting became the primary purpose of the board. This also validated a culture of cronyism with PSO favoring his friend becoming the Chairman.

2013 - BK taking over. This moved focus away from execution and put heavy focus on things like DEI, GPTW etc. and rapid decline commenced. Intel poured billions into x86 handhelds market but couldn't come up with a single suitable product. This resulted in tablets with Intel being sold for scrap prices on Alibaba.

2015: ACT - Destroyed the technical talent further. Those who were good at 'managing' their career survived, Those who put their heads down and did the work lost out. There has been a marked decline in execution since then.

I never understood the anxiety of Intel to acquire the companies they partner with - e.g. Altera, Wind River, ending up destroying the shareholder value.

I never managed my career enough to move up the ladder. But I think I was a typical employee who could see that the mistakes were being made but yet kept the trust in the company.

What finally drove me out was the continuous execution issues. Not a single product or process was released on time in the last few years as the company continued to bleed money and talent. To recover, Intel needs a series of very competent decisions and actions, zero mistakes and a strong vision. As things stand, I doubt if Intel has the right people at the top to make it happen.

by
| 3893 views | | 22 replies (last July 16) | Reply
Post ID: @OP+1k07cc9z2

22 replies (most recent on top)

Quoting OP: "SET: Bain or someone advised PSO to separate people management & functional ownership. This decision looked innocuous at that time but has had the most devastating effect on Intel."
Hear Hear.
When people used to ask me how it is to work in Intel, I used to respond that Intel was going on like a car with the best engine in the world, driven by a drunk driver. (I recently stopped responding that way because of all the good talent we lost to others)
The fact that a role called "TA" (Technical Assistant) for managers even exists, is a verbal admission that managers don't have the required technical knowledge. It's right there in the name!!! And all the TAs that I met, didn't have the required technical knowledge either.

by
| | Reply
Post ID: @by+1k07cc9z2

@bp OP might be saying that he was a fab tech and never moved beyond the Gr54 at which he was hired.

That would explain a lot about his apparent delusionary state and general lack of grip on reality.

by
| | Reply
Post ID: @bq+1k07cc9z2

Stupid Fake Troll Post.

There is no such thing as a L0 employee and everyone responding is either really gullible or the OP.

by
| | Reply
Post ID: @bp+1k07cc9z2

Totally right about the DEI hiring, even at the green badge module tech level. The temp service had an in person meeting with all the S4 green badges, and each person was asked to introduce him/herself and tell the group a little about themselves...a big group of pink and blue haired nose ringed, inappropriately dressed kids who would state their name and then run on in detail about their particular s-xuality. As if this was the first thing to know about them in a professional work environment. Out of a large group there were only about 8 actual adults who looked professional and didn't prattle on about what kind of people they like to engage in in--------e with. That's what DEI hiring gets you. You end up with the island of misfit toys.

by
| | Reply
Post ID: @bg+1k07cc9z2

@av I had a similar experience with hiring under BK.

by
| | Reply
Post ID: @b7+1k07cc9z2

My daughter in college is studying the demise of Intel in her business management class. Intel failed because of hubris, questionable DEI hiring practices, poor technical judgment and failed execution over multiple CEO’s. I gave her my two cents on 10nm failure. Sohail and his “yes men” like Charvat convinced BK that SAQP would save the day and the Co/Ru for lower interconnects was production ready. It was not and had inherent reliability issues.

by
| | Reply
Post ID: @b6+1k07cc9z2

WTF is "L0 employee"

by
| | Reply
Post ID: @b2+1k07cc9z2

I don't believe OP works at Intel.

by
| | Reply
Post ID: @aw+1k07cc9z2

Good post. To add to it I really feel during BK’s reign of te---r we really lost site of attracting and retaining great engineers. I worked with HR during that time and we were solely focused on hitting the diversity goals no matter what. We literally turned away great non diverse engineers in huge numbers to hire a diverse candidate. We used to make offers to diverse candidates at the universities without even interviewing them, just so we could hit the outlandish numbers of diverse hires. Then we spent a good part of the focal budget to reward them. It was a good cause and we needed to make sure we had a diverse pool, however we went way to far and it impacted the good engineering talent we lost. The other piece
Of this was finance. They had too much influence and they cut future money making businesses because it was not making the margins right off. Finance is what will be tagged as the demise of Intel. you

by
| | Reply
Post ID: @av+1k07cc9z2

What’s L0 employee?

by
| | Reply
Post ID: @at+1k07cc9z2

@aq I thought OP was making a long-winded, mendacious point about something, but then I fell asleep so not really sure what he was droning on about.

by
| | Reply
Post ID: @ar+1k07cc9z2

@ap I don't mind the trolling, as pointless as it may be, but am bothered by the fact that so much coal had to be burned in a power plant for the internet to carry the load of OPs 10,000 word treatise on blahdeblah.

Think about the children.

by
| | Reply
Post ID: @aq+1k07cc9z2

@OP Great, another mind-numbingly long Cut & Paste post from the AI Troll.

by
| | Reply
Post ID: @ap+1k07cc9z2

The stickiness of the x86 SW ecosystem was predicated in decision making at "modern Intel" and reinforced by partners (cough cough MSFT). The implosion of Itanium (or really any non-x86 effort) resulted in a "you don't fck with the recipe" mindset that involved a 5% IPC CAGR and lots of new ISA features on x86 line (Ts).

That strategy worked well enough during an era with no competition (~2007 to ~2017) but the world was shifting. During that era

  • mobile became a thing and created a SW ecosystem around ARM
  • web became the platform of choice for anything outside gaming
  • AMD delivered a competent design with Zen (patent sharing agreement from anticompetitive business was a godsend to AMD)
  • CPU competition was able to offer at least 50% better tcomp and had growth opportunities
  • Companies Intel supplied were able to build more aggressive microarchitectures than Intel (Apple)

Once that happened, 5% IPC (maybe) + the *T dejour (yippee DRM in secure enclaves or whatever) started to look pretty unappealing. Add in technology delays and the wheels really fell off. Add in quality issues for more even more pain.

by
| | Reply
Post ID: @am+1k07cc9z2

Great post OP. Let me simplify. Intel doesn't give a rats azz about you unless you're part of the "click".

by
| | Reply
Post ID: @aj+1k07cc9z2

@a9 , understanding past mistakes is key to learn from them. OP has done a good job with that. It is up to the management to accept and learn from the mistakes.

by
| | Reply
Post ID: @ab+1k07cc9z2

Here is how Intel learns from past:

  • When 13nm got delayed, we were told there were learnings which will be beneficial to for the subsequent processes.
  • Come 10nm, they again made aggressive and risky choices, leading to endless delays again.
  • It went on, e.g 5N4Y
by
| | Reply
Post ID: @aa+1k07cc9z2

OP, I hope your post made you feel better. You put a lot of effort into it.

I think this post and a lot of others like it on this site exemplify Intel's real problem. Intel and its employees are fixated on the past but ironically they never learn from the past. How Intel got here doesn't really matter now, especially if they aren't going to learn from past mistakes. What is important is where it goes from here. Instead of focusing on the past, focus on the present and plan for the future.

by
| | Reply
Post ID: @a9+1k07cc9z2

I really think the attitude that "everything must be IA" really hurt Intel. The fact that we were so stubborn to think Atom could compete with ARM in mobile phone space was a multi-decade mistake. Imagine if Intel had kept StrongARM/XScale and built out that platform and competed with the likes of Qualcomm, Samsung, MediaTek, etc. The current miss on AI wouldn't seem as dire if Intel had a healthy share of the smartphone business.

by
| | Reply
Post ID: @a8+1k07cc9z2

@OP beautifully said

by
| | Reply
Post ID: @a7+1k07cc9z2

after considering the acquisitions, i have to wonder if intel was used as a slush fund with kickbacks going on

by
| | Reply
Post ID: @a6+1k07cc9z2

This is summed up really well, Altera is the latest in a long line of acquisitions that they buy expensive and sell cheap. Other companies can do this cough AMD and ATI why can't Intel?

I remember in one Q&A years ago someone said that Intel is opposed to risk taking. But that is not true. The exec said that every time they push for a new process node it is risking the whole company.

The real problem is execution. AMD can execute, Nvidia can Execute, TSMC can execute. Why can't Intel? If execs are not able to figure out this problem and fix it then Intel has no future.

by
| | Reply
Post ID: @a5+1k07cc9z2

Post a reply

: