Intel plans thousands of job cuts in face of PC slowdown
bloomberg.com
bloomberg.com
Intel certainly has its own unique challenges, but even, for example, AMD had been doing pretty great up until this year, and they also just announced a big shortfall for their 3rd quarter results. Just wondering if this is really the tip of the iceberg for a true, broad retrenchment in tech, after the past 9 months of "Is a recession coming? Is a recession coming?"
A recent laptop I purchased, as well as the last desktop I put together (~2 years ago) each have Ryzen chips. I forget the details but in addition to performance issues, didn't Intel CPUs also have some major security vulnerabilities? And was it that they were related to instruction-level performance optimizations that, when disabled to address the security vulnerabilities, led to even worse performance?
So if AMD isn't doing great at the moment either, I can't imagine how hard Intel has been hit. I don't know anyone who is buying or recommending Intel CPUs at the moment.
Aren’t the major cloud providers still mainly running on and buying Intel?
https://blog.cloudflare.com/designing-edge-servers-with-arm-...
Not sure about Azure.
Meltdown (Spectre v3)
Spectre v3a
LazyFPU
TLBleed
Spectre v1.2
L1TF/Foreshadow
SPOILER
SpectreRSB
MDS attacks (ZombieLoad, Fallout, RIDL)
SWAPGS
For example, there have been demos of spectre in-browser, but applies to multi-user environments or simply apps that you didn't grant admin privs...
If you treat your computer security in a DOS/Win95 fashion, then none of them are relevant.
Tho overall having a more stable PC that consumes less electricity is better in long term and ppl really see that.
It shows in sale numbers.
As a result, the Intel processors have TDPs and real world power usages >2.5x that of a comparable Ryzen. Sure, they're winning, but at what cost? The 12900K at 240 watts pulls almost the same power as a 280W 64-core Threadripper.
AMD is responding in kind, with new top-end processors pulling 170W, or higher with their built-in overclocking that pushes the chip to even higher power draws as long as cooling permits. This looks to put them back into the lead, but it's just not a sustainable strategy.
Anyone without solar panels?
With these ludicrous power requirements, you may in fact need to rethink how you use your power. E.g. if on same as a non-heat pump drier, you'd have to make a choice between gaming and drying clothes.
Having said that I saw a reference to a 4090 offering 95% of the perf at 60% of the power usage if undervolted, so that becomes an attractive option now.
I absolutely love my 5800X3D's insanely low power usage (insane = performance per watt for gaming in simulator titles where it runs circles around Intel).
Don't get me wrong, the 5800x3d is a phenomenal CPU. However like many owners of that CPU, I'm also in the market for a 4090 and intend to use the full 600W power limit with my water loop and outdoor radiator. CPU power consumption is just not an issue for enthusiasts.
So maybe configure that as an optional profile for titles not needing maximum juice?
I'm quite keen myself.
Power $ mentioned was peak times in Norway; I'm in Australia where it's not anywhere that bad yet. (0.25 AUD for me).
That's neat, but 5% is a lot when you're spending $1600. Oh and all these measurements you see tossed around are done with air coolers and stock voltages. Of course more power doesn't help if you're limited by thermals. Specifically, you're limited by hotspots you can't see because they're between the on-die temperature sensors.
> maybe configure that as an optional profile for titles not needing maximum juice?
4K gsync monitors are typically limited to 120Hz. If my GPU is capable of running a title at 1200fps, it will idle 90% of the time and consume only ~60W regardless of the 600W power limit. So tuning this is pointless. I'd still get 1/10th the rendering latency compared to a GPU that's only capable of 120fps. Same logic applies to power limiting your GPU which will increase latency and kill frame pacing in exchange for a tiny reduction in electricity bills.
This completely misunderstands heat transfer. If the hotspot temp is 75 deg even with overclocking you're not limited by thermals: https://www.youtube.com/watch?v=zc-zwQMV8-s
>4K gsync monitors are typically limited to 120Hz. If my GPU is capable of running a title at 1200fps, it will idle 90% of the time and consume only ~60W regardless of the 600W power limit. So tuning this is pointless. I'd still get 1/10th the rendering latency compared to a GPU that's only capable of 120fps. Same logic applies to power limiting your GPU which will increase latency and kill frame pacing in exchange for a tiny reduction in electricity bills.
This completely misunderstands power consumption and the nonlinear relationship between power and clockspeed. Performing the same work in the same time in short bursts of high clocks and high voltage uses more power than constant low clocks and voltage.
That's very dismissive of the entire EU, which has lots of enthusiasts, and these aforementioned expensive electricity prices.
Some electricity providers in Germany even raised their prices to 1.35€/kWh (most famously DB Netz Bahnstrom plans)
[0]: (Search for "cumulative amount of energy required to perform Blender and x264 and x265 HandBrake workloads") https://www.tomshardware.com/reviews/intel-core-i9-12900k-an...
[1]: https://arstechnica.com/gadgets/2022/09/intels-first-13th-ge...
Hence power draw is meaningful for desktop computers, not just energy usage for a task.
AMD market cap: $93B
Intel market cap: $104B
And now you will no longer be able to say that everything you've heard indicates that AMD is "absolutely crushing Intel".
Their multiplier means that the market thinks highly of their ability to continue to grow.
*edit added YTD for clarity
Intel market cap in 2015: $150B
Kinda looks like crushing.
But 2be fair amd is/was growing and intel is stagnant and lose market share right now but with a massive investments and going back on track with fabs process i would assume Intel soon start 2grow again (ofc after the end of the recession)
We have to remember AMD is making money out of TSMC advantage over Intel node which I assume won't last forever and if TSMC blunder even one node it can be catastrophic for AMD. Considering USA shift and focus on tech war with China Intel fabs can only grow faster or everything can crash.
Within the past 12 months:
Peloton laid off over 4000. Snap 1280 (which is 20% of the company,) Shopify 1000 (10%.) Groupon laid off 15%. Salesforce, 1000. Microsoft 1800. Carvana (while not "big established" it's still a lot of people) laid off 2500. Tencent laying off 5500. Alibaba: 9500, ByteDance: 1000, Zillow 2300.
This definitely isn't really the first major layoff announcement.
Outside of big tech: Credit Suisse laying off 5000, Ford 8000, Telefonica 2700. Societe General 3700
A bunch more, but these are the one measured in thousands or otherwise a significant percentage of a company's workforce.
(By the way, not arguing with you, my point is that this isn't surprising, the writing has been on the wall for the past year, so an Intel layoff isn't a bellwether for things getting bad -- it's a lagging indicator of things already being bad.
https://www.theverge.com/2022/8/10/23299499/microsoft-layoff...
I guess my main point was that, even with recent layoffs, feels like most of those folks wouldn't have had much difficulty getting snapped up by other companies, especially in engineering (not saying it wasn't disruptive to those involved). But once you start laying off 20,000 here and 10,000 there, you get to the musical chairs point where some folks are going to be left without a chair for some time.
Yeah. It certainly was not a surprise and the market euphoria was going to all end in tears. Saw that a mile away several months before it happened. [0]
What folks don't appreciate is that AMD is on a roll. A clear roadmap and processes in place to execute like clockwork in tandem with TSMC. Intel can easily bounce back in about 3-4 years if they fix their foundry issues.
“America invented the semiconductor, but today produces about 10 percent of the world’s supply—and none of the most advanced chips. Instead, we rely on East Asia for 75 percent of global production. The CHIPS and Science Act will unlock hundreds of billions more in private sector semiconductor investment across the country, including production essential to national defense and critical sectors.”
https://www.whitehouse.gov/briefing-room/statements-releases...
https://news.ycombinator.com/item?id=33012137
Spoiler alert: no really good answers
The companies range from startups to enterprise, at practically all stages of funding, but they each told me their funding never materialized or investors gave management ultimatums.
A few were in crypto, two are in fintech, a few more in various b2b tech companies. But they all gave almost identical explanations: management sees a rough economy ahead and tightened their belts accordingly.
I’m not trying to be doom and gloom, I still have a job, but even my partner, who is also a software engineer, just survived a round of lay-offs at their fintech that happened yesterday. They laid off 25% of their company across the board.
They were “tightening their belt” a profitable company doesn’t have to worry about that.
My CTO said specifically “we need everyone we have and we aren’t going to be successful by laying off people. We have a vision and we need you all to help execute it”. They were bought out less than 9 months later for 10x revenue. I had moved on to $BigTech by then after leading their “cloud modernization” efforts.
The profitable tech companies are using it as an excuse to get rid of dead weight. No one is going to come after Cook, Jassy or Nadella for short term revenue misses.
Facebook and Google still have founders who own more than 50% of the voting shares. No one can come after them.
You can say a lot about the big 5. But none of them can be accused of short sightedness.
Besides, if you work for any tech company, you should be able to throw your resume in the aid and get another job. Even if you are a blue collar worker, there are plenty of companies looking. I have a friend who works in finance for a major manufacturer. He said the company had to be a lot more lax about firing factory workers because they already had a shortage.
Too bad, and so sad, that they’ll be just as eager for talent when talent inevitably moves on and finds they’re happier where they landed when they were being used as stock price pawns.
I'm not sure if I can take this as a harbinger of things to come, and I'm saying this while generally being pessimistic of the economy, I think Intel is a deeply troubled business and some deep layoffs were long time coming because Intel was spread real thin across so many businesses.
That's complete nonsense.
"The Automotive sector leads all industries in job cuts this year"
"Retailers led in job cut announcements in September"
"Technology companies followed"
Source: https://www.challengergray.com/blog/job-cuts-surge-46-in-sep...
Congrats to Apple and AMD for having done better
Microsoft ignoring AMD in their latest Surface lineup is a hint for the people looking for answers
For those of us who are bad at subtext, could you elaborate further? What are the questions that this is the answer to?
Really at least a fifth of your layoff should be managers. If you lay off about six reports, that’s one less manager you need. So 1/7. If you lay off six managers, that’s one skip level you don’t need. 1 + 6 + 36 = 43. Go to three levels and that’s 1 + 6 + 36 + 216 = 259, or 43 / 259 or about 1/6.
But if you want to keep production capacity, you should move to slightly more reports per manager, not maintain steady state. Just to keep the math simple, if you had 216 developers reporting to an organization of 43 managers, if you went to 7 reports per manager you need 31 line managers, not 36, which is a 2% RIF without losing a single producer. Or if you do both you get about 1/5th of laid off people as managers if you lay off less than half your developers.
Edit: fractions and ratios are not interchangeable, bad math.
Then 216 => 43 becomes 203 => 29
Now you've reduced the workforce by 10% and put some miserable projects entirely out of their misery at the same time!
We had 294 line managers, 49 managers, and 8+ program managers. 351 managers, 2115 employees total. I need to lay off at least 15% or 318 people.
We go from 294 line managers to 252, plus reductions up the chain to 36 + 5 = 293. 58 down, 260 to go. I lay off 5 managers, 35 line managers, and 245 staff, that’s 58 + 285 = 343 people total, so I can argue for keeping a couple extra line managers and all their reports. Call it 327 layoffs, 96 of them managers, 28.4%.
I think the entire new surface line-up is a giant conspiracy.
The Surface Pro 9 5G has a Qualcomm 8cx Gen 3 with some tweaks called the SQ3. It does not match the performance of 12 Gen intel in a 15w power envelope, which is important because the Surface Pro 9 non-5G (intel) uses an i5-1235u, a 15w chip. But the last Gen, the pro 8 used a 28w 11th gen part. This means that the y/y perfromance is actually about the same (but better power/heat). If they used a matching 28w CPU it would have been a much bigger jump comapred to the ARM SQ3 5G.
But then, it gets WAY weirder.
The new (FOUR THOUNSDAND DOLLAR) Surface Studio 2+ ? 11th gen laptop CPU (11370H)...which is as fast as SQ3 will be.
Surface Laptop 5? Exact same specs as Surface Pro 8.
Casual reminder that the now two year old Apple M1 still smokes every single chip I just mentioned, all in a fanless iPad Air. The would-be Surface Laptop AMD Ryzen 6800U would be right on par with M1...which would make that SQ3 look like a bad deal.
Microsoft is sandbagging the lineup for ARM.
Surface pads do not seem to be too popular and might become even less now
E.g. when the first Surface Pro ARM was released (the X, 2020ish), it was ridiculously more expensive, less performing and had less battery life (!) than the cheaper Intel options https://www.notebookcheck.net/Microsoft-Surface-Pro-X-Review...
The previous Surface non-Pro ARM attempts (RT, 2012ish) were even crappier (Tegra procs) but at least they were cheaper than the Pro/Intel variants...
I've got an SQ1 and Apple M2 and they're both truly excellent.
(disclaimer: I worked years in engineering in both Qualcomm and Apple)
Essentially I run them all day with Visual Studio Code and vim sessions, Firefox scanning news and documentation (and of course normal email etc), and ssh to cloud hosts.
Your mileage and evaluation could of course vary.
Well yes, but then you have to use macOS and the locked down close ecosystem that is Apple. That's not really something they're directly competing against.
The second part of your question is key though "why most people should care." They obviously don't. It "just works" and generally keeps them from doing insecure shit. Want to buy hardware from someone else? Tough shit! Why would you anyway? You've got money to burn and no desire to write code that runs on your box without a second "real" computer. Buy some more lightning cables while you're at it. Don't forget to mention green bubbles next time you message an Android peasant.
I said that last part pretty snarky, but you're not wrong about most people not caring. That's their audience and they've nailed it.
The conversation never mentioned “iOS” or phones and the submission is about Intel who has nothing to do with mobile or Android. He specifically said MacOS:
>> Well yes, but then you have to use macOS and the locked down close ecosystem that is Apple.
What does any of your response have to with Intel, Windows or MacOS?
>> Casual reminder that the now two year old Apple M1 still smokes every single chip I just mentioned, all in a fanless iPad Air.
Far too long managers are not taking responsibility for their bad decisions - not only business decisions but also simple feedback disregard.
„Look at what happened to Intel thats what happens if you let your managers loose”
When I was there, they called Intel "Great place to leetcode" (a corruption of Intel claiming they are a "great place to work" everywhere)
Intel is known for good WLB and quiet quitting because the non technical managers didn't have a clue about what was actually being done on their own team. It's easy for engineers to BS these kinds of people, and this was the norm since any passion for SW was killed by working there long term...
Yeah, its always them that get cut \s
One touched a lot of teams including my own. The second was mostly smoke and mirrors - they took multiple factories and supply chain employees and sold them or outsourced them to a company that picked up the payroll for those employees. Only a handful of roles were truly culled.
The announcement of the layoffs was theatre for Wall Street.
This is due to the New York Stock Exchange being located on Wall Street in New York.
I'm interpreting this as "the (publicly-traded) company used a layoff announcement to appease investors and show them that executive management is taking steps to ensure the company operates more efficiently, thereby keeping it within the good graces of these investors/hedge-fund managers/etc".
Just saying Wall Street sounds very dismissive of the obligations that management has to the owners.
1) Completely missed the boat on mobile. Their ARM-competitive chips (Atom, etc.)... weren't. Missed on every measure from power to performance. Let Qualcomm and Samsung eat their lunch.
2) Missed too many ticks and replaced them with tocks. Fell far behind because of ridiculous tilt-at-windmill folly around Itanium and other architectural decisions that absolutely didn't pay off. Ended up taking way too long to get further down on process size.
3) Lost their competitive edge so much that Apple finally realized they can do it better and suffer the transitional costs of building a middleware layer to translate off of x86. Apple's fully committed to that, so Intel doesn't just lose the "no one is using a PC anymore" market, they're also losing the "those people who DO use home computers which happen to be Apples" market.
It's absolutely incredible to watch these formerly phenomenally innovative companies falter at that very core competency of staying cutting edge but we see it happen over and over again.
To be frank, I think that's a great thing. I don't want companies to stay giant in perpetuity, I think it's great that they are essentially "recycled" when new, better companies come along and eat their lunch.
If anything I think it's quite bad that over the past 25ish years that a lot of big companies have really learned the lessons of "disruption" and have responded by just buying up the smaller but up-and-coming competition before they can overtake (looking at you Adobe and Figma).
It's bad for a company to go bad.
It's bad for a company to lock out competition.
Hard disagree, so I wonder what your basis for this bullet is.
Agree on the other two though.
1. Tautological: from company's perspective, it's bad for the company (itself) to go bad.
2. More broadly, it's reasonably self-evident that it's bad for employees, and bad for shareholders of a company, for that company to go bad
3. Where we can have discussions, and probably for a long and fun if not necessarily fruitful time, is whether it's "good" or "bad" from market's, or consumer's perspective for companies to go bad. I personally think not.
I think we'll likely mostly agree that it's good for market to filter out and punish bad companies or companies that go bad, once they go bad, for whatever reason; but that doesn't necessarily imply or follow that's it's good for companies to go bad.
In other words, what's your perspective/bias - why would it be GOOD for a company to go bad? Why would it not be absolutely fantastically wonderful if all companies perpetually stayed good and we lived in utopia of rainbows and unicorns? :>
(all this without defining what "company going bad" means, left as an exercise for the student :)
Generally where my head was at. The other two points were better scoped. I personally think it's good that companies go bad, as it moves control over productive assets from one group to another. This is good for society for a variety of reasons, not the least of which is enabling new ideas to be tried. Can a long lived company try new ideas? sure, but the degree to which they avoid risks (if they live a long time they necessarily avoid undo risk) puts on upper bound on how ambitious they are with new things.
Sometimes on net things are good because of other effects like "ah, promising new competitor gets a chance to shine!" but if you could have _both_, you absolutely _would_. A company going bad is a _cost paid_ for new blood, not an _added benefit_.
It's not good that you have 100 fewer hours when you spend 100 hours creating something awesome. It's bad that you have 100 fewer hours. If you could create the same awesome thing and still have those 100 hours to do something else, that'd be fantastic.
All other things being equal, I would far rather Intel be awesome than not. If it turns out the world is such that all other things can't be equal, it might be worth having Intel not be awesome in favor of other benefits, but I'm never going to be _happy_ about Intel not being awesome.
When they can't it's good companies come and go, or Kodak would still be blocking digital cameras (https://www.reuters.com/article/us-kodak-bankruptcy-idUSTRE8...) but the price we pay is things like wasted effort on re-implementing everything that didn't need to change at the new place, good people can't always move to it as you mention, etc.
Different countries at different times adopt different positions on the scale between inefficient (?) state-linked monopolies with jobs for life and letting the market do its thing. I'm not sure what factor means that sometimes we end up with Samsung and other times British Leyland.
That itself is an example of capitalism as the least worst option... similarly how much effort goes into trading currencies or commodities or whatever just so people get a fair price and aren't screwed over by whoever is the only person selling at the moment they need something. But we're self-interested, biased, and this is the workaround (or the system that emerges from our nature - self-interest is turned into the energy behind it all and "greed is good").
Don't you have on the order of tens of thousands of dollars invested in Intel through retirement savings, though? When S&P 500 companies fail, your retirement savings become worthless. (Apple is really the one you want to watch out for, though!)
I don't love this; the total stock market is a lot of tech companies, and I already have plenty of exposure to tech by working in the field. It's a good heuristic for most people, though. (And no, I don't do anything about this underlying fear of tech sector exposure. I just buy the Target Date 20XX funds like everyone else.)
My only point is that the grandparent comment generally expects companies to die when they get into the S&P 500. If that's true, we're all screwed. If one S&P 500 is just stealing business from some other S&P 500 company, though, it's probably a net gain. But if it's some privately owned startup, then it's not as concrete a win, and let's be honest, startups are driving a lot of the innovation in tech.
If I had a time machine, I would definitely use it to time the market!
My savings account is not what I'm talking about. Inflation or not, it is still a number I can describe as "positive".
I'm referring to bonds that retirement and wealth management add to funds use to supposedly counter volatility. These are investments that at best will only pay single digits. They have suffered the worst year since 1931 or something. And we're just getting going. The frustrating part is as I said, the underlying bonds themselves still pay that 4% or so which would be much better than the funds are achieving. But for some reason they have to sell the bonds at a loss.
fund managers don't have "personal" or "corporate" assets that are anywhere near the scale of the (retirement) funds they manage; being held responsible for mismanagement would not lead to a source of capital to replace that which had been lost through mismanagement.
All investment has risk.
It seems the person upthread owns a target date fund and this means they probably are diversified. Their post doesn't make a lick of sense.
"When S&P 500 companies fail, your retirement savings become worthless. (Apple is really the one you want to watch out for, though!)"
If a major player is massively overvalued, though, then index funds, too, will feel some pain as the exuberance in a major player dissipates.
I hold some INTC directly after having been extremely impressed with the company's dedication to rigorous process in engineering interviews in the mid-2010s. I'm sad to see such a great company go through hard times.
Why is my fund in Intel but not in Apple? The whole point of choosing the index is so that you are diversified. You would only lose if Intel lost out to a foreign company, otherwise the demise of Intel is made palatable by the rise of Apple and AMD.
Companies come and go from it all the time:
https://www.inc.com/ilan-mochari/innosight-sp-500-new-compan...
"Unless you bought inte directly, the rewighting should handle it. Unless it fails quickly and spectacularly"
I would find little comfort in the fact that if Intel goes down 85% -slowly but somewhat spectacularly- from its current level it will eventually be excluded from the index.
The fact that I had to google for the details of its fall gives me comfort that when intels time comes it won’t cause harm on its slow exit.
Except that semiconductors are a national security issue at the moment.
Maybe too big is a national security issue, too.
Too big to fail is a real problem for any part of our economy.
Perhaps, or perhaps we're dealing with monopolization.
Plus, dictatorships arent against the US per se, they re against the US convincing the populace they too could act like americans, and are extremely dependent on that Schrodinger state where you re both the factory for the US and a public political opponent. Symbiotic parasite pretending to be the host's alternative.
Plus the US might change heads at the top, but it s hardly a safe ally to have. Ukraine nearly went to complete disaster thanks to Trump and having morons elected there is a security risk for many too...
Perhaps this is too much of a radical idea, but if a company is so critically important that it can't be allowed to falter without government support, then it should be nationalized.
At the very least the government support should be made explicit, and large, critically important companies should be forced to pay much higher insurance, i.e. the equivalent of what banks need to pay for FDIC insurance. Sick of the whole "privatize the profits and socialize the losses" mindset.
An investor would need to front many billions of dollars for an uncertain payoff over a decade or two, and Intel could at any moment become competent again. And if no investor is willing to take that risk, there would be no new competitor.
Adobe and Figma are software companies, so don't require capital investment. So if one shuts down, it's relatively easy to make a new UI modeling tool(or for an existing one to take over their customers).
It's a disastrous process, I passionately hate it
I think their main error was short-termism. They were addicted to high margin (workstation/server) devices. They had no interest in the low margins of mobile devices. They could not imagine how low-power devices would take over the world.
Only a few companies can buck the Innovator's Dilemma, usually only founder led enterprises with significant control.
The Innovator's Dilemma isn't about missing quarterly guidance. The dilemma is that it's perfectly rational and profit maximizing to continue focusing on your cash cow even when obsolescence is a foregone conclusion. It wouldn't be a true dilemma, otherwise.
The incumbent is the only player that can maximally squeeze the very considerable remaining profits from old technology, and they should do so with gusto. Moreover, switching to new technologies comes with more risk, even when it seems obvious what the new market will look like because the old market has almost zero risk--it's completely proven.
All the "solutions" to avoid the dilemma, like selling the old technology to take future profits and then pivoting to the new market, are just corporate branding shell games. They might even be in fact sub-optimal, but in any event the fundamental dynamics remain the same.
There are transaction costs to creating and building a corporation, but do those offset the clear costs of leaving money on the table, especially in the modern world of highly liquid capital, and particularly in markets with clear technological breaks. The lesson of the Innovator's Dilemma is that very often the perfectly, unqualifiedly rational decision is to press your advantage to the very end. And importantly it not only maximizes short-term profits, but implicitly it maximizes long-term profits globally by most efficiently allocating resources. Why waste energy swimming upstream when there are endless fish spawning and starting their journey upstream already along with ample resources of their own.
This is reported in a profile of Otellini:
> "We ended up not winning it or passing on it, depending on how you want to view it. And the world would have been a lot different if we'd done it," Otellini told me in a two-hour conversation during his last month at Intel. "The thing you have to remember is that this was before the iPhone was introduced and no one knew what the iPhone would do... At the end of the day, there was a chip that they were interested in that they wanted to pay a certain price for and not a nickel more and that price was below our forecasted cost. I couldn't see it. It wasn't one of these things you can make up on volume. And in hindsight, the forecasted cost was wrong and the volume was 100x what anyone thought."
https://www.theatlantic.com/technology/archive/2013/05/paul-...
So a deal based on any price wasn't a realistic avenue for iPhone. The fact that Intel was actually considered was itself a radical move on behalf of Steve (absent the technical obstacles that emerged later).
Is there any evidence that mobile chip makers who do not also create/own the device and are in a race to the bottom on price actually get decent returns?
No. Managing innovation at the scale of Intel is a challenge. And their customer base (PCs and servers) is shrinking. Also, this is only like 15% of the workforce. Which is only 5% more than a typical culling.
He and his whole team were furious.
IBM crippled the AS/400 when it looked like it could eat into their mainframe business, for example.
It's the innovator's dilemma.
I think of it as a peer to the 2005 proposition that electric cars didn't need to be golf carts and the technology was coming into place to make high performance, real cars that were purely electric. Could have been true. As they say "very dangerous, you go first." That one paid off.
I wonder if there is a list of large dollar, novel, engineering expeditions and how they turned out.
␄
¹ Where "miracle" means some really smart people worked really hard at it for a long time.
Rather Intel wanted to leave the baggage of x86 behind, segment the market for higher profits and hoped to get better performance than superscalar x86. But they never achieved the latter and most people just wanted 64 bit and software compatibility.
The first superscalar x86 processor was the Pentium, which came out in checks notes 1993 (Wikipedia claims design work started in 1989, hit working simulation in 1990 and taped out in 1992). Intel didn't start work on Itanium until 1994, after Pentiums were already shipping to customers, and Itanium itself wouldn't ship until 2001.
And, it may be paying off after all. Graphics cards are effectively giant VLIW machines. And while Intel's new cards get slaughtered on DirectX 11 or older games, they are quite competitive when used on the DirectX 12+ stuff which basically wants a giant multi-core parallel processor.
I'm gonna add the caveat that the failure didn't happen out of the gate. Xscale was leagues ahead of anything else two decades ago. Atom stumbled at the start but Cherry Trail had a better price to performance ratio than anything that followed it. Intel's failure was nothing but pure avarice.
My impression is Intel hit the price for embedded just fine, but at higher watts, and thus they can't compete.
That was also excellent for Cherry Trail, with an SDP of 2 watts. When you have some time sit down and compare the benchmarks from a Cherry Trail tablet like the Surface 3 and its successor chips like the 4415Y. The 4415Y like the one in the Surface Go is three years newer, has a 60% higher TDP, a Recommended Customer Price over four times higher, and while its 3D chops are better it actually benchmarks lower than the Cherry Trail chip in PCMark.
As someone who worked related to those projects (PowerVR, which supplied graphics IP for some of those devices, on Android so other OSs may have a different view), I am completely not surprised they didn't go anywhere.
This was a good few years ago - I no longer work for PowerVR, and Intel seem to have completely given up on the market for some time. My memory isn't likely prefect, but I can give broad strokes.
It always felt like "Having" to go to PowerVR was an embarrassment for the teams, they kept trying to replace us with their internal GPU architecture, presumably completely fail to hit any power/performance targets, then last second call us again and try to rush everything through. Then, it feels like most of the time they drop the entire project before it made release anyway.
The teams that we worked with never felt high status - all the engineers we spoke to were either on that team because they couldn't move internally to something more prestigious due to some internal politics, or were actively in the process of changing teams. There were times where our internal engineering contacts changed monthly, if we had any engineering-level communication at all.
There's also some weird public claims about things around this - like PowerVR not providing driver source or similar, but that's simply not true. As the person packaging up the releases, they were only source releases, no obfuscation, with full documentation and build instructions.
The PowerVR driver model generally meant releasing a "reference" driver as source, with hooks for customers to hook up to their specific SoC implementation (think stuff like setting clocks, power management, bus endpoints etc.). The supplied package only had a couple of example backends. While one of those may have been an Intel chip, it was only minimally setup in order to make the graphics work for PowerVR testing. It was simply not possible to release a binary-only driver. But I've seen this claimed on a number of tech forums - often from people claiming to have knowledge from the Intel side, and if that's true and not exaggeration, I can only assume it's due to lack of communication between teams internal to Intel. I can assure you, someone in Intel had the full driver source.
Though WRT the internal communications, I sometimes saw this from the other side - we dealt with 2 Intel teams with different SoCs they intended to build, and it felt like 2 different companies. They never shared anything between them, almost as if they never spoke at all, and seemed resistant to changing that. I have no idea why, I guess it's just how their management structure is setup?
If you go through all that and end up with an actually good product, I feel it'll be almost luck rather than anything else.
I applied to Intel's architecture group in 2015, and when I asked one interviewer "does Intel have a strategy if TSMC catches up on process tech?" I was left with a blank stare, as though nobody in the group had contemplated the possibility. I didn't end up working there...
That said, Intel got exceptionally unlucky on which technologies they pursued and which they didn't, so I hope they can turn things around. They missed on through-silicon vias and chiplets, EUV, Cobalt wires, and several low-power transistor technologies.
Do you have a source for this? The only one I found says Intel is 2x TSMC:
https://www.statista.com/statistics/883715/microprocessor-ma...
By wafers produced (including memory), TSMC is 2nd and Intel is 6th. TSMC is 2x the 5th placed company on the table, so Intel is even smaller than that. https://www.eetimes.com/chipmakers-increase-share-of-global-...
Wafers at 14nm or better would probably be a good metric for who has the volume behind top nodes to then be able to stay in front. If that's TSMC by a large margin losing the mobile and GPU markets has really done a bigger number on Intel than I had assumed before.
"It means that computer purchasing decisions are no longer made based on price/performance or just performance, like in the dark ages. Now, when somebody decides to buy a new computer it will be price alone, or maybe price and service, that determines which computer to buy."
[0] https://jlforrest.wordpress.com/2015/12/18/the-forrest-curve...
https://www.tradingview.com/chart/?symbol=INTC
It's one of a few large companies with that distinction.
Shares now offer a (relatively) regal 5.3% dividend yield and a price/earnings ratio just above 5. Of course, as profits dwindle, both metrics will be recalibrated downward.
This seems relevant because during the late 1990s Intel was allegedly the company whose shares one bought and never sold. I could name a couple of those whose positions today are argued to be equally ironclad.
As you observe, this is true only if one ignores dividends, which are a material component of total returns for a mature company like Intel.
> during the late 1990s Intel was allegedly the company whose shares one bought and never sold
This has never been true for any public company. And it’s driven by investor style more than company fundamentals.
Not really true. Most billionaires utilize the "buy, borrow, die" strategy of borrowing against their holdings instead of selling to get around taxes. As long as the equity increases in price more than inflation you come out ahead, plus save the 20% on taxes. Then when they die their capital gains are reset so no taxes ever paid.
https://totalrealreturns.com/s/VFINX,VBMFX,USDOLLAR,INTC?sta...
Because the rest went bankrupt.
https://stockcharts.com/freecharts/perf.php?INTC
The picture is better, but not great. Folks who bought at the peak of the bubble would have needed to hold for about 18 years to break even.
US treasuries are nearing 4% and are risk free.
And your dividend % is before tax, so compare to tax-free bonds.
Is it just me, or do some other folks also smell "corporate bloat, which better management would not have allowed to happen"?
If you're a Linux user, Intel integrated graphics have always been the most compatible and most well-supported option.
Speaking on AMD, 10 years ago at-least, there were very few premium AMD laptops, and they used to overheat quite a bit, has that changed?
GPUs of the era your thinking of had high failure rates from issues with lead free solder, though the Nvidia GPUs on Macs and Laptops would outright fail from other issues, requiring a full chip replacement.
https://eclecticlight.co/2015/12/20/lead-free-graphics-cards...
For the Linux users I was referring to, the most graphically intensive thing many of them run is a desktop compositor.
But yes, today, AMD iGPUs are a great choice for mobile, and AMD dGPUs are a great choice for desktop or for i-don't-care-about-battery gaming laptops.
On an unrelated note, "perverts" doesn't seem to me like a particularly kind appellation.
You forgot to add, "worst performing by a mile" to the feature list.
Of course I can install the proprietary one, but that means I can no longer just upgrade the system and expect it to work after reboot.
I get your point, but intel video options perform on par or better with regards to the most common consumer video rendering demands at this point. Video acceleration and high resolutions and multiple displays are well covered -- not everyone needs to process GPGPU workloads and play the newest games at 90FPS.
I mean, it coincided with that. It's not clear how much was caused by branding/marketing and how much was caused by Intel being better (at least in laptops) from the Core Duo days until now[1] plus-or-minus a few years?
---
1. Honestly, I haven't kept up with laptop hardware performance during the AMD chiplet era.
the pentiums were a huge leap from the amd 386/486 and cyrix options. 'Intel Inside' was basically a premium-product differentiator for those that could afford it, and that was well understood by consumers at the time.
Ferrari/Gucci/Armani/Rolex/Louis Roederer labels also help to push product. Same phenomenon, people didn't buy Intel strictly because it was needed for specific workloads, they bought it because of the fancy sticker that differentiated their product from cheaper alternatives; even if the person didn't know thing-one about computers or CPUs.
It sounds corny, but having lived through it I can vouch that things are really that stupid.
I always buy Intel. I simply do not care for AMD's software jank, Intel has consistently proven more stable and reliable and that's not something I'm willing to trade for marginal differences in performance caps.
With the rise of mobile gaming, the fact that people don't know what chip is in their phone casts even more doubt on the value of brand awareness. Intel got dominant by always having a design and/or fab dominance over rivals. A "one-two punch." Who among PC buyers understood that?
On top of that, if Gelsinger is serious about building a contract fab business, that has sales and marketing needs way outside of anything Intel does today.
If you’re at a company and you look at your engineer to sales ratio you might think to yourself “gee, we have so many sales people!”
That’s because there’s not much multiplying factor to sales roles. There isn’t a lot of potential to automate like engineers can do. A salesperson only has so many hours in the day and they’re fitting maximum one or two customers into a one hour slot of their time.
Might be different for Intel though, everyone has to drink at the same oasis.
People will believe any statistic that aligns with their mindset.
Unemployment rates are at historical lows within a small margin, it's been 50 years since unemployment has been lower (i.e. something like 60% of Americans have not seen lower unemployment in their whole lives), and since the WWII war economy that they've been significantly lower (i.e. almost nobody alive).
In the UK people sometimes make the claim that the unemployment rate the government is quoting is so low because people have given up even looking
(unemployment rate) = (people not currently working, but actively looking) / (total people)
The first includes all people, the second excludes those not working, but also not looking.
"The Australian Bureau of Statistics (ABS) defines a person who is unemployed as one who, during a specified reference period, is not employed for one hour or more, is actively seeking work, and is currently available for work"
So if you are desperately looking for more work BUT you have a casual job providing 1 hour of work THEN you are NOT unemployed. I mean, it's right there that you have a job.
Given the high rates of casual/part time work in our workforce, unemployment rates mean nothing to me anymore...
https://tradingeconomics.com/euro-area/labor-force-participa...
The rest of intel is a shambles. The networking group in particular is an industry joke.
This article extrapolates that out to "All divisions will cut 20% of staff".
It's unfortunate for a lot of people regardless, but the report seems (for now) to be exaggerated.
Maybe if Apple's products were still "Intel Inside" Intel wouldn't be in this position.
In that case, the deeper read of the situation is that the Covid crisis helped them kick the can down the road by 3 years on responding to the trends of 2019.
Piss poor management strikes again.
You could kinda say they made the PC? Also involved in PowerPC processors.
Not getting of mainframes were not their problem I think.
By contrast IBM made boatloads of money off mainframes and their advantage was durable. That business line is still quite profitable even today. (Just not very big in comparison to the IT market.)
[0] https://www.forbes.com/sites/timbajarin/2021/08/25/attack-of...
I think Intel just bought into their own Wintel uber alles bullshit and couldn't even conceptualize devices with TDPs under a watt. No one could possibly do anything worthwhile without Wintel so shed everything that's not Wintel!
I have little faith in Intel's management. They apparently plan to continue increasing their dividend payouts, while cutting back on investment in fabs according to https://semianalysis.substack.com/p/intel-cuts-fab-buildout-...
... and in the context of the recently-passed CHIPS Act...
I wish they'd be more of an engineering-driven bottom-up company. IMO there's a huge company culture to contend with that is primarily top-down. So much scar tissue in the company prevents changing things for the better. So many good people capable of making these sorts of changes get burned out and leave the company. Intel's compensation compared to FAANG certainly helps to push people out of the company.
Good thing Pat is back and now going with EUV to remain competitive.
With all the things we book under the "National Security" banner a much needed next gen chip factory should be one of them.
“The factory of the future will have only two employees, a man and a dog. The man will be there to feed the dog. The dog will be there to keep the man from touching the equipment” – Warren G. Bennis.
Also Intel never had a hyper-inflated share price like most other semis. It's actually quite comical just how much the share price has dropped because its previously meagre dividend yield is now 10% due to denominator effects.
The buy back would similarly help ex-employees with stock or options, and I'm sure the package on the way out from Intel will be quite good.
See: https://www.tradingview.com/chart/?symbol=INTC 50% down YTD.
I am pretty sure there are lots of bloats at Intel that requires trimming down.
Just had a little flashback to 2008 me
Gamers (and non-Mac high-end desktop performance users) would seem to be a pretty small slice of the pie these days. I have a high-end current M1 laptop but I do a lot of my work on a couple of 7 year old Intel-based desktop/laptop machines which are perfectly fine although they're about to go out of OS support. (Though I'll probably use them for a few more years anyway.)
I remember when you wanted a new PC with every tick of the processor cycle. These days, for most people, who cares?
Interesting situation. I never touched a PC until after college. But, yeah, there was at least a period where basic "computer literacy" was expected. Maybe things have shifted again.
I acknowledge I'm in a unique situation where I run a home lab and so I have a local cloud, but because of that, I can effectively do 100% of my job off an iPad Pro (I don't want to for quite a few reasons, but I could). If I didn't have my home lab, I could get close by deploying my OS on AWS/Digital Ocean/etc. and have the same effect (though latency would drive me nuts eventually + internet connectivity issues - though this would break my job anyway).
If I weren't coding, I'd probably entirely switch to the iPad. I do all my document editing on it. I do my code reviews on it, architecture planning, etc.
I could realistically see the average consumer never needing anything more powerful than something like an iPad Air. I complpetely believe we will head to the era where the only hardware the average consumer needs includes:
* Great resolution screen * Great internet connection * Great input controls * Minimal CPU/RAM to run a robust, hardened portal to cloud services
iPads are basically beefier versions of even that.
I'm happy using a good Chromebook for most things but I've never found an iPad--even with external keypad--to be an adequate substitute for anything involving text editing.
I'm fairly happy with the magic keyboard with it, but understand different preferences!
Interestingly, Mac sales are up 40% recently. So I guess a lot of people actually do care -- they just want in on the best laptop processors out there, which are decidedly not offered by Intel.
With no good Windows support for ARM, I'm curious what trends we'll see over the next few years for PCs. Will they languor in their current state, using Intel's hot n' heavy processors? Will some manufacturers switch to ARM + Linux solutions to offer battery life and heat generation competitive with the Mac?
>Will some manufacturers switch to ARM + Linux solutions to offer battery life and heat generation competitive with the Mac?
I don't know. You already have Chromebooks but those aren't really mainstream outside of education even though they're all a lot of people need. And Google exited as basically one of the most high-end hardware makers there.
Both Qualcomm and Samsung produce tablet chips that perform well enough. I wonder why they haven't expanded into the laptop space? I imagine Apple's M-series would be hard to compete with, but many consumers would be fine with something competitive with the M1, which seems doable.
https://www.notebookcheck.net/Qualcomm-Snapdragon-8-Gen-1-Pr...
It's like we've all gone back to the time before free software and GCC. It's a bit worrying that the PC market is shrinking. The PC is the only freely programmable computer on the market today...
[1] https://www.gartner.com/en/newsroom/press-releases/2022-10-1...
Intel's dominance was practically guaranteed for five years, I knew Zen was going to be successful but I was surprised to see that they've successfully flipped the script and Intel is in a bind.
How to tell an article is written by a hedge fund shorting the company.
Are you referring to past leadership? Because Pat Gelsinger has never been a CFO...
He was never a CFO.
this will age poorly.
Intel is already fighting back strongly and really only had 1-2 iterations where they weren't in some ways the performance kings. They'll be back, fabbing their own chips, while AMD and others pay TSMC to make them, and intel will make massive profits.
Intel 12th and 13th gen competes with AMD on performance and price and its still using yet another 10nm finfet process while AMD is using TSMC's latest whatever.
none of this is indicative of a 'finished' organization. It's just not quick, because nothing in basic research and chip design is quick.
The metrics that matter financially are performance/power and performance/area, and Intel is worst-in-class in both metrics in both CPU and GPU right now.
I worked in chip design at Intel for over a decade. In the 2016 culling, I noticed they laid off a ton of smart people, but all the terrible management and fake-it-til-you-make-it engineers survived. I left very soon afterwards. I suspect the 2022 massacre will be along the same lines. Intel as an organization is not just finished, it is terminally toxic and incapable of being fixed.
That sounds like what happened to Boeing.
The why is more interesting to me and seems pretty self-evident: a CFO's job is to keep bad things from happening.
That's not the sole person you want running a company, because that's just dying more slowly, with good numbers.
No one really worked and would go home early.
It was really weird.
I also have a 14 year old laptop. That one's days are probably numbered. Maybe I'll get one in the next two years.
Increasing productivity is a good thing actually.
a company that has low profit and many employees is a lot more beneficial for society than the opposite
He who steals the fires from the gods, can force his standards upon men for the time the flames last.
Around 30%-40% of PC sales were going to cloud compute. This is slowing.
On one hand, the data center gets better utilization, maybe gets more value, and maybe turns over hardware faster. (e.g. why do I want to buy a new computer when the IGPU is just going to make it crash faster?)
On the other hand there is more competition for the data center, particularly cloud providers who could amortize rewriting simple but large scale applications like Amazon S3 for ARM or RISC-V over a large fleet of machines. If the data center is able to drive a hard bargain, it may well be that the client is still subsidizing the data center, but we just get told its the other way around so that we won't ask for me and complain about the e-waste Intel tries to pass off on us. (e.g. "try" because their sales are collapsing)
---
A good example of the gaslighting is this article
https://www.tomshardware.com/news/linux-kernel-update-kills-...
which should have the headline "Intel iGPU kills laptop displays".
The company developing a competent alternative to the M1/M2 or nvidia GPU is going to have a good business.
Seems to me that Intel is alway in this cycle of greatness and then crash while sitting on their laurels. It seem they need to have a fire under their buts to make the right decisions.
That is to say, i9 13900K will have 8 performance cores and 16 efficiency cores if I recall correctly. Many of the mobile CPUs in 12th gen also have more efficiency than performance cores.
Listen to interview with Jim Keller, yes, there is a whole team doing the work, but you need a technical/management person with the credibility and willingness to make the hard choices and take risks.
99% of people hired for CEO roles will be happy to coast, take their bonus on short time gains and move to the next opportunity. Who is willing to take risks for long term success that will not appear on the quarterly report?
You need someone who can get that job and at the same time doesn’t give a damn about what people think and about the money. Those people are rare. That’s why Apple doesn’t have ’another’ Steve Job, Apple will coast on the previous trajectory for a while, but won’t disrupt anything else major, they will play safe.
The same is true for Google, they got their own money printer going nonstop, who would be crazy enough to take risks with that?
The last disruption of Google was Gmail/GSuite and was started as a side project. How many top engineer do they have, where is all their output?
For Intel they need to disrupt their own products to move forward, that’s why it’s hard to do, you don’t want to kill the golden goose. But it’s either you now or the competition in a few years.
Man Intel was really sleeping on this one to be kicked that hard by Apple on the first release of their desktop chip.
So in a nutshell: talented engineers are everywhere in those companies, that’s not the diferenting factor, good leadership is and the incentives reward those who don’t take risks.
BTW: your comment is mostly an ad hominem with a condescending tone: mind of the laymen, incredibly lazy thinking, I love comment like this!
Try with arguments instead.
Your original post had no argument, either, so I don't see the necessity.
Now that you've provided an argument, we can have a discussion, starting with:
> Intel was really sleeping on this one to be kicked that hard by Apple on the first release of their desktop chip
This isn't true. The M1 in a desktop isn't competitive against Intel chips.
Intel can't execute. They are failing to shrink transistor size. Failed at a 5G modem. Failed to deliver a competitive mobile processor.
I don’t think a company like Intel was ever designed around innovation past what made them dominant in Windows PCs. And like Microsoft, they’re going to need to diversify.
there's something I don't understand: car companies and everyone else can't seem to get enough Chips and yet Intel is laying off employees?
car companies need specific chips for their cars. Intel makes its money selling different chips. Car companies don't want to redesign for different chips all the time.
https://www.gartner.com/en/newsroom/press-releases/2022-10-1...
The people you need to grow fab and manufacturing capacity are different than the people you need to design cutting-edge new products. And the sales team you need to offer fab services to other companies is different than selling CPUs to consumers.
1. Is desktop computing still relevant?
2. Is server CPU market not making enough return?
3. I thought Intel needs talent to bootstrap their US-based fabs.
4. Does this mark the beginning of an end for PC?
They get what they asked for: EFI, ME, expensive processors, feature locking, core as a service the bullshit that USB is.
But it doesn't look good for Intel on that front either.
Original text:
It is. The article does not say where this will occur.
They cannot lay off people from the EU like that. It's not that easy. Unlike the US, EU countries are welfare states.
This is the case in Denmark, some countries offer incentives to lay staff off temporarily but, generally, it isn't the case (which is why you have unbelievable levels of unemployment in Europe).
In Germany, there is a thing called "Kündigungsschutz"[1].
I am not sure about Ireland, but I assumed there are similar rulings.
So at least in Germany, it is not that easy, but possible.
1) http://www.rechtslexikon.net/d/kuendigungsschutz/kuendigungs...
Either way, the point is that if you have high levels of protection that should mean that it is easier to fire someone...because they have something to fall back on. That is why the UK and the US have universal welfare states. Germany does not have a universal welfare state, and the only way that is possible is by having high employment security/strong labour laws and huge companies that are profitable due to low competition. This, obviously, comes with downsides (this is why Germany has struggled with high unemployment in the past, and has things like Kurzabeit to subsidise companies even further against firing staff).
An exception to this is Denmark which has almost total union participation and very weak labour laws, they achieve this through a much more expensive system of social insurance. They have, for Europe, relatively high levels of innovation so this all offsets (you get high security and relatively high levels of innovation).
However, I know for sure that in Germany, it isn't that easy to hire and fire people on a whim. In the US this seems to be the case.
Just in case, let me also clarify this:
I don't care, if you think the American system is better than the German one. I simply made the statement that even if not all EU countries have rulings like Germany, it is surely harder to fire people on a whim.
And also the article didn't mention where this will occur.
And one particular aspect of Germany is that capacity is controlled on the way up. The reason why the US has these huge swings is because capacity isn't controlled. It isn't possible to have the upside without the downside. Germany's political economy is totally different because the US has a far higher level of competition and innovation, this isn't possible with Germany's labour market.
AMD for a long time had been headquartered in Sunnyvale, CA, but they moved just a few years ago. They are now building homes at the former AMD site.
Just remember, we're not in a recession.
I still stand by what I said about already being in a recession months ago since November, in this classic comment: [0] but then one said:
"No. That is completely false. We are most certainly not in a recession. Nor did one start in November. You can check the Fed data here and verify you claim is false"
"The economy is considered to have entered into a recession if it experiences a decline in GDP for two consecutive quarters. And that has not happened."
So those in [1] still waiting for the official figures by the Fed, NBER, etc which those are lagging indicators to tell them it is too late? I don't think businesses would like the sound of that.
Really unsurprising, that all of this is happening. The time to prepare was November.
??? Hasn't there been booms and bust cycles in america since founding?