Intel seems to be bleeding out ...

[DiTBho]

No. All modern electronics lifetime equals to guarantee period. So you can expect that it will be broken just after guarantee period expiration. And it really fails just when guarantee period expires.

This also happens in the world of bikes 

The simplest example is the inner tube: the modern ones in latex - 30 bloody euros per inner tube - are offered with a 2000km guarantee, and you can be sure that at 2000+1km the valve will pop and they will have to be thrown away and buy again.

Luckily, compared to computers, you just have to not buy them, and use the good old technology of butylene inner tubes - 8 euro per inner tube - which are still made ...

... if only there was this possibility with computers

[DiTBho]

technology to precisely predict lifetime for a short period

Weibull distribution + research and development conducted by Google when they were using cheap computers and components and had to figure out  when to replace components before a failure occurred, this to minimize both the expense and the downtime.

They did a lot of research on continuous probability distributions

[coppice]

This is paranoid nonsense.

No. All modern electronics lifetime equals to guarantee period. So you can expect that it will be broken just after guarantee period expiration. And it really fails just when guarantee period expires. Old electronics usually has much more lifetime and worked much longer than guarantee period. This is because there was no technology to precisely predict lifetime for a short period. Modern technology allows to use materials with predictable lifetime.

For example, modern mixer uses plastic gears inside which wear out and fails after 1-2 years (guarantee period). Old mixers were made with metal gears and can works for years with no fail. This is how programmable lifetime works...

Of course there is some of this going on. I suspect a lot of the motivation to get mains voltages out of many products, and confine them to a wall wart, was so they could use rapidly degrading plastics to house the main product. I find it suspicious that when people like the EU started bitching about plastics not degrading in landfills we suddenly had a lot of product cases becoming brittle and break easily not long out of a fairly short warranty. Surprisingly, the wall warts don't do this, as that would cause manufacturers safety liability issues.

That said, most of the manufacturers I have worked with set life targets which aren't that awful. Like most TV makers seem to target 8 years of trouble free operation, which isn't too horrible a target. Many places will throw in a 5 or 6 year warranty as a sweetener, even if the manufacturer's warranty is just one year, because failure rates up to 5 or 6 years just aren't that bad, unless the designers messed up on a particular model.

[David Hess]

Every single Macintosh computer I have from the 80s, 90s, or 00s that I didn't sell when I upgraded (or that I bought used cheap later when they were obsolete) still works.

All of the PCs that I built for myself going back to the Pentium II Celeron 300A still work.  The Pentiums that I build immediately before that failed because the embedded lithium batteries in the clock and CMOS chips ran dead.

[brucehoult]

Every single Macintosh computer I have from the 80s, 90s, or 00s that I didn't sell when I upgraded (or that I bought used cheap later when they were obsolete) still works.

All of the PCs that I built for myself going back to the Pentium II Celeron 300A still work.  The Pentiums that I build immediately before that failed because the embedded lithium batteries in the clock and CMOS chips ran dead.

So we are in agreement that the vast majority of consumer products (especially electronics) don't just die for no reason after a couple of years.

[David Hess]

Every single Macintosh computer I have from the 80s, 90s, or 00s that I didn't sell when I upgraded (or that I bought used cheap later when they were obsolete) still works.

All of the PCs that I built for myself going back to the Pentium II Celeron 300A still work.  The Pentiums that I build immediately before that failed because the embedded lithium batteries in the clock and CMOS chips ran dead.

So we are in agreement that the vast majority of consumer products (especially electronics) don't just die for no reason after a couple of years.

There are some exceptions, even with my still working PCs.  Incidentally, all of these PCs have ECC memory, which is not something commonly found in consumer products.  I never had memory problems, but if I had, then it would have been detected.  The one PC that I did not build, a Fujitsu laptop, died a couple months after the warranty expired, and this was AFTER two motherboard replacements.

Some of them have been through 3 or 4 power supplies.  I have had a couple of Corsair and EVGA power supplies which just spontaneously quit, but this was within their warranty period.  I had a bunch of identical Vantec power supplies which all failed at about the same time not long after their warranty expired because their output capacitors dried out.  I am sure that a careful manufacturer could design a product to reliably fail after a given number of hours of operation by tuning the operating life of the aluminum electrolytic capacitors.

Fans regularly wear out.

Every nVidia card I had quit after its warranty expired.  Just one of many Radeon cards failed outside of its warranty, but most never had problems but became unusable with operating system updates.  One Radeon card had a poorly designed cooling solution causing it to fail after a couple years, but with a lifetime warranty, XFX replaced it with a better one.

[radiolistener]

So we are in agreement that the vast majority of consumer products (especially electronics) don't just die for no reason after a couple of years.

Chinese guys can made chips which can dead just after several month of relaxed usage... I have no idea how they doing it, but it just stops to work...

It's entirely possible that Intel is introducing a new technology that makes the life of chips even shorter

[djacobow]

Regarding self-destructing chips....

When I was a wee engineering student, we designed chips and had them manufactured on these free MOSIS "tinychip" runs. Later, at real fabless chip companies, it was possible to get your design added to a low-cost "shuttle" run.

In both cases, the parts came back unpassivated. They worked for a few weeks/months, then died. Good enough for a student, good enough for testing a design, but when you want to make some chips to sell, you good pay for your own very expensive mask set.

[floobydust]

Intel plans to cut thousands of jobs to finance a "recovery". Stock -40% this year. Yes, bleeding out.
The 13th and 14th gen CPU crashing and failures, they're still for sale... Everyone is jumping over to AMD for now.

[David Hess]

It's entirely possible that Intel is introducing a new technology that makes the life of chips even shorter

Everything is consistent with a design error, like with the Atom processor failures around 2017, or a process related problem.

My recollection is that in the past, Intel high performance processors were designed for a 20 year operating life, chiefly limited by temperature, which led to operating temperatures of about 70C at most.  Many processors are now operating at 90C which makes me wonder if they are designed for a 5 year operating life.

[mengfei]

Intel plans to cut thousands of jobs to finance a "recovery". Stock -40% this year. Yes, bleeding out.
The 13th and 14th gen CPU crashing and failures, they're still for sale... Everyone is jumping over to AMD for now.

dang! how are those around 15K goon cope up this with this kind of economy
Never thought this will happen to the CPU leader

https://www.theverge.com/2024/8/1/24210656/intel-is-laying-off-over-10000-employees-and-will-cut-10-billion-in-costs

[madires]

And more news:
- Legal eagles target Intel for class action over cooked Raptor Lake CPUs (https://www.theregister.com/2024/08/01/intel_raptor_lake_lawsuit/)
- Intel announces two extra years of warranty amid chip crashing and instability issues — longer warranty applies to 13th- and 14th-Gen Core processors (https://www.tomshardware.com/pc-components/cpus/intel-announces-an-extra-two-years-of-warranty-for-its-chips-amid-crashing-and-instability-issues-longer-warranty-applies-to-13th-and-14th-gen-core-processors)

Two more years of warranty won't cut it.

[coppice]

Intel plans to cut thousands of jobs to finance a "recovery". Stock -40% this year. Yes, bleeding out.
The 13th and 14th gen CPU crashing and failures, they're still for sale... Everyone is jumping over to AMD for now.

dang! how are those around 15K goon cope up this with this kind of economy
Never thought this will happen to the CPU leader

https://www.theverge.com/2024/8/1/24210656/intel-is-laying-off-over-10000-employees-and-will-cut-10-billion-in-costs

So, you know nothing about Intel's history.

[iMo]

https://abingtonlaw.com/class-action/consumer-protection/Intel-Processor-Issues-class-action-lawsuit.html

[SiliconWizard]

So yes, that's looking really bad. I wonder what they call "non-essential work". Did they take some advice from Boeing? Maybe.

[floobydust]

It's going to be 15% or 15,000 people gone to save the target $10B the memo of doom and more source.

Intel will do what all losing corporations do - fire as many engineers as possible, then replace them with salesmen 
and pay the salesmen in wage AND bonus as much as the engineers earned lol. Many such cases.

[DiTBho]

It's early August, and there's very serious news: Intel has lost 35% of its stock value, and is about to lay off 20,000 people

[DiTBho]

15,000 people

actually they are 20,000 for 2026.
not to mention the crappy way they interact with journalists

[JPortici]

The new fab in italy?



what about the other plants they were planning to build in europe?

[SiliconWizard]

Good question. They may have flagged them as "non-essential" too.

[iMo]

It could be easily their cpus architecture, silicon process and technology, company organization, company processes, amount of "none essentials", etc. reached a point pushing Intel into a steep "decline" phase, or something the shareholders would name "a loss of momentum", etc.

It happens in all large companies after several decades of a grow, a success, when the company and its products have passed the phase of "maturity" (see the picture).. The Maturity phase for Intel could be the 25 years from 1990 to 2015, imho..

Btw., 125000 employees w-w is really too much for what Intel is doing, imho. They will go down to some 40000-60000 within next 5 years, my bet (I worked at 4 large w-w corps, 3 of them did the same decline, the 4th will do as well)..

PS: in the graph you may see that typically during the maturity phase the cash/revenue holds, but the profit declines - that is because the companies usually hire more and more employees (usually sales and marketing, management and other none-essentials), and/or they start to dig into the biz areas which are not their core business..

PPS: FYI - AMD almost tripled the number of their employees since 2016 (to 26000 as of the end of 2023), thus having today 5x less employees with 2x more revenue per employee compared to Intel.. 

[Bicurico]

While some seem to be amused with what is happening to Intel (and wish them the worst), I am very worried, instead.

What are real alternatives to Intel Core i7 and Core i9 desktop processors?

AMD Ryzen? What if you need to run professional CAD/CAM/CAE software that was only certified for Intel processors?

Can one be sure that i.e. simulation will produce the 100% same result if run on an AMD CPU (not certified) as opposed to an Intel CPU? We had problems in the past in this regard...

My Core i9-13900KF is a powerhorse. The workstation runs 24/7 (not constantly on load, of course) and it is used for NC toolpath calculation and NC machine simulation, among other tasks. The speed is incredible compared to former Intel CPU's and the fact that there are plenty of cores, means that I can run several tasks in parallel.

What if my CPU dies? What are my options? Another Core i9? Switch motherboard and move to AMD Ryzen? Will that be compatible with all the software I need to run?

What about all our customers who all use Intel CPU's with their CAD/CAM/CAE software and will call us because the software crashed...

I sincerly hope that Intel gets their act together, fixes the manufacturing issues and provides propper replacement to all customers without much of a fuzz. It will be expensive for them, but do they have an option?

[Monkeh]

AMD Ryzen? What if you need to run professional CAD/CAM/CAE software that was only certified for Intel processors?

Can you specify software with that limitation?

Can one be sure that i.e. simulation will produce the 100% same result if run on an AMD CPU (not certified) as opposed to an Intel CPU? We had problems in the past in this regard...

I would immediately abandon any such software as untrustworthy, myself. That should not occur unless they're playing dirty tricks with undocumented and therefore unpredictable features.

[harerod]

...
What if my CPU dies? What are my options?
...

If this is a business machine with a socketed CPU, you are in a rather comfortable position. Should you encounter BSOD before the updates from Intel reach your machine, you simply swap the CPU and continue working. Surely the money for the CPU is nothing compared to lost time? So maybe just buy a spare CPU in advance? Then file for a refund by Intel, should the BSOD happen.
The only PITA could be software licensing, node locked to the CPU's serial number. If you are running actively supported software, this shouldn't be too bad, though.
In comparision: I have a complete set of spare components for my i7-4770k workstation. However, losing the CPU would most likely trigger issues with several older permanent Adobe and Microsoft licenses. Since that CPU is plenty fast for the software that I am running on it, I already throttled the peak power to make that machine last. (if I understand the issue correctly, this would not work for the current CPU's with the faulty microcode)

[Whales]

While some seem to be amused with what is happening to Intel (and wish them the worst), I am very worried, instead.

What are real alternatives to Intel Core i7 and Core i9 desktop processors?

AMD Ryzen? What if you need to run professional CAD/CAM/CAE software that was only certified for Intel processors?

Can one be sure that i.e. simulation will produce the 100% same result if run on an AMD CPU (not certified) as opposed to an Intel CPU? We had problems in the past in this regard...

My Core i9-13900KF is a powerhorse. The workstation runs 24/7 (not constantly on load, of course) and it is used for NC toolpath calculation and NC machine simulation, among other tasks. The speed is incredible compared to former Intel CPU's and the fact that there are plenty of cores, means that I can run several tasks in parallel.

What if my CPU dies? What are my options? Another Core i9? Switch motherboard and move to AMD Ryzen? Will that be compatible with all the software I need to run?

What about all our customers who all use Intel CPU's with their CAD/CAM/CAE software and will call us because the software crashed...

I sincerly hope that Intel gets their act together, fixes the manufacturing issues and provides propper replacement to all customers without much of a fuzz. It will be expensive for them, but do they have an option?

Both hardware and software need to do their best to be compatible.  Intel and AMD do a fine job.  Your software vendor sounds suspicious if they're whitelisting CPUs.  Worst case their software depends on timing quirks (race conditions) of certain CPUs and they don't want to fix it.  Best case they simply have not bothered testing on other CPUs due to cultural limitations in their company.

Onus is on the software company to fix this problem, not you.  If they don't "certify" new stuff or fix the bugs that prevent their software being compatible then they'll be scaring and pissing off their users.  The users should not be expected to fix this problem unless they are given a way to test "certify" a CPU themselves.