Intel pays more, can attract superior talent.
Maybe the Indian H1B is just a fraud.
29 replies (most recent on top)
@OP here’s why businesses (like Intel) have issues, according to entrepreneur Alex Hormozi
“Most founders think pushing harder and moving faster is the key to growth.
I've seen the opposite.
The businesses most obsessed with "growth hacking" and "scaling fast" often grow the slowest - because they get stuck fighting fires (for years).
It looks like a plateau but feels like drowning.
Here's why:
Rushing creates short-term thinking, which creates long-term problems. Think about it.
When you rush:
You hire fast (and don't train or onboard). You skip systems to "move quick."
You patch problems instead of fixing them. You make snap choices you later regret.
You launch a quick new service or product without adding infrastructure - you "double time" a team that's already spread thin.
It doesn’t work. Believe me, I’ve tried.
The math is simple:
Every "shortcut" you take today creates two new problems tomorrow.
Those problems each create two more.
Soon you're drowning in issues that all trace back to one thing “a rush” - an arbitrary goal that you set for yourself that you said if we do not hit this I will be unhappy.
And in trying to pursue it, you guarantee your failure.”
No infinite OOP. They work more than 5 hours a week. Or a month in intels case.
Intel has no way to compete with TSMC. Intel will fall behind further.
Reasons:
- The top 3 managements are too incompetent, they are just talkers without technical knowledge.
- "Cheating" is the way to survive in Intel. People who tell truth and facts have been chopped off.
- No innovation. Intel's copy-exactly policy has trained people to copy and follow. They lost technology throughout the years. After the rounds of layoffs, talents have left.
- No virtue and no moral. This is sadly a place darker than most place on earth. Politics and dishonesty dominate without any punishment. The cheating managers can do whatever they want.
@ef
Those managers do not contribute decent work and
they are way overpaid.
How did they get their positions?
This is a top-down problem.
The rot starts with the board and works its way down.
Intel is not technology-oriented. The managers cheat all the time to get credit for themselves. The cheating ripples up and down and all the scheduling/planning/vision in Intel is distorted. Too many people cheats that turns facts upside down. Some managers at upper level does not know anything, but they cheat or say words so vague to blur the mud.
The failure of Intel was years in the making. Excessive share buybacks that should have funded R&D or acquisitions, selling of ARM XScale license to Marvell and failure in mobile, no compelling AI product, continued loss of client and data center market share to AMD. In the end the X86 moat failed and Intel had no coherent strategy.
The Far East have embraced broad state guidance of key monopolies while the US has had a haphazard industrial policies at best, letting companies like Intel, Boeing, etc pi-s away long term prospects for short term gains.
@b9 yes 18a is marginally better than n2 on some metrics and will come out 3-6 months before n2 products:
But in very low volumes because Intel is the only customer and they can’t achieve the economies of scale to recover the costs.
A sort of pointless victory.
@b8 you could argue that Intel’s reluctance to go with EUV was a cost issue given the level of PC and server units. In fact Intel was very early to EUV R&D with significant investments in ASML (cool tech that they could not bring to market).
The economics didn’t make sense at the time for the stagnant PC market. For all of the sh!t Intel takes, their engineers are not generally stupid and they probably made a reasonable choice to pass on EUV.
The economic did make sense for smartphones (obviously as TSMC showed) and they pulled ahead, accelerating their virtuous cycle.
@b8 Are you trying to say 18A is better than N2? Intel 3 is better than N3E? Intel 7 is better than N7?
You cannot bring them to market because they su-k. Go on, lie to yourself. Your customers are not coming unless orange man forces them to
@b7 as I mentioned before.
You can have great tech, but if you can’t bring it to market due to a lack of units (to justify capex) it’s moot.
Winning smartphones would have provided the units and the lead would have been maintained.
Intel failed to diversify products from PCs and servers into other higher growth areas.
@b6 Look, CapEx affects how much you can manufacture only, R&D affects how good your tech should be. Intel is behind in both. Stop beating around the bush, face the issue.
@b5 it’s not R&D, it’s CapEx.
You need to amortize the cost of the fabs across many units.
Intel doesn’t have the units.
Of course Intel has some extremely wasteful and ridiculous R&D spending. That’s a separate issue, but Intel can’t afford leading edge fabs without going in the red.
@b4 Apparently Indian frauds cannot read financial statements, so I will give it to you here. It has nothing to do with income, it is purely to do with how much is spent on R&D.
Here are some R&D figures (US$ MM) for year 2022 / 2023 / 2024:
Intel - 17528 / 16046 / 16546
AMD - 5005 / 5872 / 6456
TSMC - 5327 / 5944 / 6223
@b2 if it’s not a money issue, explain the billions in losses, and chronic negative free cash flow and the government bailout?
Someone making the statement that it’s not about the money is not a serious person.
Sure you can develop some cool technology in the lab, but bringing it up to HVM scale requires big bucks.
Can you imagine Tim Cook saying - sorry the iPhone is delayed because Intel fvcked up 14A?
Or Jen Hsun telling his customers that Rubin has slipped because of 14A.
I mean. It’s ridiculous on the face of it.
Ain’t nobody going to risk literally 10s of billions in revenue on Intel. Even if Intel did have some marginal tech lead, TSMC would follow up with the same tech in a much safer way (like BSPD) very shortly afterwards and customers would wait because it’s a safer bet.
OP summarized it well. They are frauds. Intel did not skim on R&D so it was not money issue
Some would say Intel is even behind Samsung.
How did Intel fumble the Tesla contract to Samsung?
I mean how???
I guess Intel as a credible foundry is just a fairy tale.
@ay also notable are the trillion dollar hyperscalers like Microsoft, Google, Amazon and Meta all building in-house silicon to power their data centers.
It’s hard to compete when your competitor has an infinite money hose.
Semiconductors are all about economy of scale and TSMC ships 10x the wafers vs Intel. Intel is sub-scale and cannot compete without these external source of funds.
Nobody is going to risk their bread and butter products on an unproven foundry with a history of failure, conflicts of interest and deteriorating financials.
It’s a viscous cycle.
It’s very simple.
Intel missed smartphones.
That’s it.
With that, Apple gave TSMC more money than god to speed past Intel.
The next massive growth wave - AI has also been captured by TSMC as Nvidia is giving them unlimited funds in parallel with Apple.
Not to mention all of the secondary players like Qualcomm, Broadcom, AMD, Mediatek etc…
Intel isn’t competing with TSMC.
They’re competing with the combined firepower of all of their very wealthy customers.
@ab nope SAQP has zero chance against EUV
That’s why Pat switched out horse
So you're telling me there might be a chance?
also Intel backed a losing horse Self Aligned Quad Patterning when TSMC recognized EUV was the future
Not sure what you are saying, friend. Intel is out front.
Playing catcher.
Intel did not attract talents at all. Most of the Master's degree candidates interviewed are Indians (h1B), and this ends up more and more Indians. It is not balanced and no talents.
I have to say, when I was a hiring manager at Intel not long ago the hiring queue was always 99% Indian and from Berkley or Merced. I often had to source my own candidates. Perhaps diversifying their sourcing pools for talent could be beneficial. I know many people who wanted jobs at Intel, but somehow never surfaced through the queues.
It’s pretty friggin crazy… just the salaries TSMC offers.
I guess discipline and hard work beat out fancy degrees and H1Bs?
Not to answer a question with a question but take Raja and Jim as superior talent attracted by Intel.
That helped Intel pull ahead, right?
Brian Krzanich, CEO from 2013 to 2018.
Technical delays: Under Krzanich, Intel sc--wed up with delays in its 10nm, which let TSMC catch surpass Intel in manufacturing technology.
AMD comeback: AMD made a strong comeback with its Ryzen and EPYC lines during Krzanich's leadership, exploiting Intel’s process delays and architectural stagnation.
Corporate focus shift: Shifted too much focus away from core CPU engineering toward trendy areas like wearables and drones, many of which didn't pan out.