Intel's processor failures: A cautionary tale of business vs engineering Just like Boeing, once upon a time, Intel was the darling of the engineering world. Both companies were the premier tech companies in their day, but those days are long gone now. As for Boeing, what can you say about an aircraft manufacturer whose planes crash and doors blow off in mid-air? Oh, and which now appears unable to …
Boeing is hardly the only one... It started late 1970'ies early 1980'ies, the neo-liberal market, and its primary driver is called Shareholder Value. Nothing matters but the shares and the value they represent. No longer about products. No more competition on quality. Just get the investors happy by increasing the "value" of the company. It is about money, nothing else matters.
And even the shares only matter until you can hype them to the sky and then flog them off to some sucker.
After that, they can - and will - crash into the ground, and those responsible not only won't suffer; they'll have made off like bandits, providing a strong incentive to do it again.
What exactly is efficient about this method of running the market?
"What exactly is efficient about this method of running the market?"
It's a terrible model but on the other hand, many of the companies that get contrasted with Intel in this piece and many others as the "heroes" - be they nVidia, AMD and so on - are operating within exactly the same national business environment.
Seems to me one of if not the key difference here is that almost all of those "hero" companies (including ARM and Apple's chip operation) are all relatively puny fabless semiconductor operations that have Big Daddy TSMC and others actually manufacturing all their stuff for them, and don't actually have to keep a state-of-the-art semiconductor manufacturing operation running around the world all by themselves the way that Intel does.
Intel may not have a competitive 5G modem chip right now (if you're comparing to Qualcomm - also a fabless company) but neither does that other giant vertically-integrated silicon manufacturer named Samsung have one either. (Just compare those Samsung-built modems in the Google Tensor SoCs for evidence) Nor does Broadcom these days, either, as far as I know. Qualcomm is in a class by themselves, if we're talking about top-tier performance.
So anyway, in reality Intel is doing a lot of heavy-lifting that all those fabless competitors don't have to bother with, and that sucks up a substantial portion of their resources.
Could they execute better? Sure. But unlike most of their so-called "competitors" they also don't have to worry about things like China invading Taiwan, taking over TSMC and UMC and shutting them down, either.
Only controversial since a couple of governments switched from recognizing the Chinese government in Tai Pei to recognizing the Chinese government in Beijing as the legitimate government of China (and sticking to the one-China policy, but as that is even in the Taiwanese constitution, no real blame to any non-Chinese government).
@Filippo
Well....this sort of market cycle has been around for a very long time.......
Other commentards say it started in the 1970's.....no.....how about the 1870's!!
Take a look at the economic history of railway development in the UK.
Between the start (around 1830) and the final consolidation, there were multiple "railway bubbles" where investors piled in........and lost their money.
The only people who made money were the final group (GWR notably)....groups who bought up built out assets for a penny in the pound....and went on to make a fortune.
Yup....bean counters have been hard at work for a VERY, VERY long time!!!!
P.S. Talking about "bubbles"......what do you think of the NVIDEA stock market valuation......or about Elon Musk's demand for "compensation"????
If you want to discuss bubbles, that goes even further back. Ever hear about the Tulip mania in the 1630s?
I don't know where my copy of Mackay has gotten off to, but fortunately it's available from Project Gutenberg, so I was able to confirm my recollection that in addition to the short chapter on Tulipmania he has rather a longer one on "popular admiration of great thieves", which might be instructive.
Of course Mackay also discusses bubbles that came after Tulipmania, notably the French "Mississippi stock" bubble — without which the USA might be a much smaller nation today — and the South Sea Company.
He also has a long chapter on the Crusades, and while he doesn't really touch on this point (IIRC), those were treated as investment vehicles (e.g. the Italian commenda) by some Europeans. (And then of course there's the failed "Crusade" of Étienne of Cloyes, the second of the so-called "Children's Crusades", which just resulted in his remaining followers sold into slavery.)
Irrationality and investment have likely gone hand-in-hand since investment became a discernible practice.
One of the biggest problems of modern management theory is that the short term measure of shareholder value is in direct conflict with the idea of a corporate personality with a goal of long term survival. It values present shareholders over future ones. Humans have a sense of long term self preservation, companies are dependant on their officers to provide that.
Examples of this are:
- There's a huge short term benefit in cutting R&D but in the long term these are injurious to the corporation's long term prospects because you fall behind your competitors.
- Similarly you have those who encourage selling off and doing a leaseback on your physical premises, which is fine till rents go up.
- Lastly you have companies borrowing large amounts to pay dividends, which is fine under low interest rates but the moment they go high you're in trouble (e.g. Thames Water)
Some of this can be priced into shares but usually they end up being rocks under the water, ready to sink the company. One way you could mitigate this is creating a duty to the long term survival of the company for the C-suite where instead of paying short term bonuses, you force longer term deferred compensation based on long term performance. Alternatively you make it the market's problem in the form of long term dividends rights which are paid 3-5 years later to whoever owned shares on a specific date in the past the right of which then would be something tradable so you then trade these on the long term prospects.
I feel you have some things right but I remember the days way back when the MD's (see how long ago) salary was altered to include share options to more align them with shareholders. Unfortunately the effect seems to have aligned them with making money for themselves with the options
<< Alternatively you make it the market's problem in the form of long term dividends rights which are paid 3-5 years later to whoever owned shares on a specific date in the past the right of which then would be something tradable so you then trade these on the long term prospects.>>
Once you "make its the market's problem" the emphasis changes to "how does the market make money out of this?" Look at the response to climate change and carbon credits etc
The result at Boeing didn't become publicly apparent in time for Intel to learn the lessons, I believe. You can argue that GEC in the UK taught the lesson well, but from the point of view of a US corporation, that would be a long time ago, in a galaxy far, far away.
Having said that, it isn't a binary choice of investors or engineers, it's always about balancing the interests. A truly excellent product that's far more expensive than one that's almost as good but much cheaper isn't going to sell at all well (look at the fate of the German camera industry). Boeing and Intel simply leaned too far towards short term investor interests.
GE in the US has to all intents and purposes ceased to exist in the form that it was created as.
The only difference between US and UK corporations is that the UK tends to be more of a pioneer, it 'financialized' and 'de-industrialized' earlier than the US. Although Margaret Thatcher was credited with this she really only finished the job, the process had been under way for decades before and only the valiant, and ultimately futile, efforts of the 1960s Labour government slowed the process.
>A truly excellent product that's far more expensive than one that's almost as good but much cheaper isn't going to sell at all well
Not necessarily. The problem is that once you get a hit product there's a tendency to just milk it, cutting forward investment so that when that product eventually gets too old the crash becomes inevitable. (The Mini in the UK is a good example -- great idea, neat implementation, should have bought breathing space for the car industry to regenerate but it didn't.) Intel actually had warning of the future but chose to ignore it -- by the 2000s the (American designed) Pentiums were just about finished but the company was effectively saved by the Israeli designed parts, originally developed for mobile, which became the processor line for the next couple of decades.
GEC was quite the outlier with its own cocktail of disfunction.
Credit control was hugely damaging - the rule was that suppliers weren't paid until they put you on stop and you needed something. So projects would grind to a halt because components hadn't turned up. The opening position with any new supplier was "we'll require six months credit".
Acquisitions would be bled dry and then shut down. I remember an early 1980s robotics company that Weinstock bought, moved from Watford to Rugby, then starved of development cash until its products were old-hat, at which point it was shut down.
The end result was the infamous cash mountain which meant that GEC became a takeover target. The good thing was that after one junk-bond financed takeover scare, Weinstock panicked and merged swathes of the company with Alstom, at which point some of the financial fuckwittery lifted.
This was all before the Harvard MBA/McKinsey types got into their stride and started hollowing out business after business. And so now we have long-view Chinese companies that make long term business decision and, when they pay off, the MBAs try and put up trade barriers rather than relearning how to run engineering businesses.
Oh, you mean that exact thing that some stupid pension fund is blaming Intel for trying to do now ?
Business over engineering always fails. Until MBAs get that drilled into their heads, there is no way companies will ever learn the lesson.
Sorry ,
I’m an engineer that worked at Inmos a long time ago . It was engineering led. Sadly the engineers were so pleased with themselves that they took pride in how hard it was to program the transputer. Only other boffins should be able to use it ….. there were many reasons for the demise, but imo this engineering and arrogance was a factor.
Hey ever checked out Valve's Handbook for New Employees floating on the internet? Apparently in a flat structure, you were suppose to go up to Lord Gabe and directly tell him why he is wrong on relevant matters and others would jump in and tell you why are are wrong instead. Fascinating read. Employees are very specifically curated and I'd be surprised if there is any cultural changes after he retires.
DEC was engineer led for a long time too, which worked when it was mainly selling to engineers. But its demise didn't come from being run by bean counters. Engineers need to work with good market researchers and others who can direct them so they don't spend too much time on things that will never sell. DEC's "marketing" was mainly well-dressed people who sat next to sales folks on sales calls and handed out brochures; they failed miserably at finding out what they should be building. Hence HP and others cleaned their clocks. Boeing and Intel are sadder stories, though, because they really did know what to build and how to build it, if the engineers were left to do it, but the bean counters took out short-term profits instead.
DEC Engineers actually did build stuff people wanted - everything from LSI-11s, MicroVAXen, (normal) PCs, Ethernet (remember DIX?) ,MIPS, and eventually Alphas. The problem was the manglement kept pushing weird old shit on customers as well - like VAX-11s (74 series TTL discrete logic in the 80s - seriously folks ?), VAX9000s, Rainbows, and VMS. Every time they'd see sense eventually - eg: shipping normal PCs, Ultrix, OSF/1, NT - but always too late and too half-arsed to make a difference. Folks also under appreciate the crushing weight of the FUD that crushed the investment, innovation and life out of any company not toeing the "No one ever got fired for buying Wintel" line.
Love him or hate him, Stallman's GCC project has opened a lot of doors over the long run - and given folks a way of the Wintel trap.
DEC built far too many things. You could have a PDP-8, a PDP-10, a PDP-11, a PDP-20 or a VAX and an almost infinite catalogue of different operating systems. And the performance differences in the various model ranges were often imperceptible because competing teams were allowed to develop similar products which marketing were left struggling to differentiate. None of it really mattered when the only competition was from mainframes, but when the microprocessor came along they were effectively doomed because they didn't have the economies of scale to sustain silicon development for their proprietary market and were entirely the wrong, ponderous, research-heavy sort of organisation to produce cutthroat commodity hardware in volume.
It's not just about having engineers or book-keepers in charge, it's about having the right people in charge for the realities of the market at the time. Or, at least, not having the wrong people.
DEC was ultimately hamstrung by the exec believing computers are big expensive things that only a few users get to use.
They made some cracking hardware and well thought out software too. But 100k/unit for ten people sharing was not going to survive a 5k PC on every desk. And even less so once competing hardware appeared in bulk.
DEC wasn't killed by management, poor engineering, or poor marketing. The company lived and prospered for a couple of decades in a niche between mainframes and devices of many sorts that did useful stuff and needed a bit of dedicated on-board digital logic to function. DEC computers were, for the most part cheaper to buy and run than mainframes and, being programmable, quite flexible. Lots of market for that. By the late 1980s it was apparent that microcomputers could, or would soon be able to do the same stuff. Cheaper. Sure enough DEC simply couldn't compete and was bought out by a Microcomputer operation (Compaq) in 1998.
I also worked at INMOS. As a PFY I found Transputers were very easy to write code for (in comparison to say PCs, 68Ks, 6502s, NS32016s, VAX-11s), whether that was in OCCAM, C or even ASM. For the record I am categorically *not* a boffin and at least the folks I worked with weren't prideful, arrogant or smug - so you're tarring them with an overly broad brush IMO. FWIW I think some of the engineers at INMOS earned the right to a bit of pride - formally verifying a FPU from paper to silicon was amazing (and paid off - judging by the hit & miss efforts of the competition at the time), coming up with a decent PLL on a digital process was a big deal too (something Intel eventually caught up some years later with the 486DX2). Also easy to forget that the power/performance ratio of INMOS gear was in another league in comparison the opposition at the time...
As you worked there, you would know that the design team for the 486 was bigger than INMOS. Yet INMOS produced cutting edge SRAMs, RAMDACs, Microprocessors, Compilers, DSPs, Video Controllers, developed FLASH & DRAM IP, and ran their own fab.
The brick wall for INMOS was securing sufficient investment to continue developing their products, which is something that all the clever engineering in the world isn't going to fix - not even a T800 with an MMU and a UNIX port would have been enough to fix the lack of investment.
TL;DR : Give the engineers a break - smug or otherwise.
Typically Transputers had four (low end Transputers had 2) 'OS' links ran at 20Mbit/sec (and you could get close to peak on that), at the time the best Ethernet you could wangle was 10 Mbit, and you got a fraction of that bandwidth - if you were lucky. FWIW the INMOS engineers did know about transmission lines - but the emphasis was on low-cost interconnect and implementing it on a process that was optimized for *digital* circuits rather than analog.
For the ill fated T9000 they implemented 'DS' links that ran at 100Mbit/sec, same deal again - they were intended to be cheap... IIRC those were being tested in the lab by 1991 - 100 Base T Ethernet turned up in 1995... They also implemented wormhole routing (that might even be another of their innovations) for the (32 port!) DS-link crossbar switch. As it happens one of the chip designers I lived with at the time explained to me about how they were using incremental power up of line driver transistors (something I saw being touted as an innovation *20* years after INMOS did it), thus shaping the signal on the transmission line and reducing the ill effect on the ground plane (eg: ground bounce).
Some of the INMOS refugees went on to promote Transputer style links, see IEEE 1355, and I heard rumors that PCIExpress was somewhat inspired by DS links. While there may be some truth in that, I figure something like DS links and wormhole routing are pretty much inevitable if you want to scale up a system and speed up comms links because Physics. It's just that INMOS tackled it in the 80s, while Intel et al took another 10-15 years to get around to PCIExpress. :)
Nah. Inmos UK wasn't 'engineering-led'.
It was driven by a vision, not by engineering.
The vision - that all systems were naturally concurrent and could be best and most efficiently executed on concurrent hardware - could well be true. But the actual implementation of the early transputers became uncompetitive fairly quickly, and the next-gen t9000 was much too late probably because it was much too complex. This complexity I'd attribute mainly to the choice of a stack-based architecture rather than a register-based one and on the lack of competent management at the UK end. But in any case, though the t800 was competitive with available RISC processors, it fell behind almost immediately, and because of choices (no cache, no page-based VM) it couldn't be used for systems which needed a UNIX-like OS. So it missed workstations initially and later the PC/Mac market.
Further, the insistence that the things be programmed in Occam was a major barrier to adoption.
Not to say there weren't good things done well - there were. The memory interface; the just give it 5v and a 5MHz clock, the 'if one isn't fast enough use more' scalability, are good examples. And the breaking graphics market provided a major opportunity.
transputers could have been killer GPUs. But alas no.
There were quite a few projects in Europe using Transputers. For example DAIMLER had an autonomously driving car in the early 1990s controlled by Transputers.
But there is no Industrial Policy in computer technology to speak of, while in the US CIA/Pentagon will ensure funding for promising startups such as Google, Facebook etc. HP was probably also a Signal Corps creation, back then.
What is needed is Airbus-style financing, developing, producing and marketing of such projects. There is plenty of scope, given Hambuger Technologies such as C, Unix and the Webbrowser/JS contraptions.
Believe it or not HP started out in the mid 1930s making electrical oscillators (some were bought by Disney for use in audio on their films) - they went on in WW2 to assist the US military in various areas around oscillators, wave analysers, shell fuzes etc. In the 1950s/1960s they moved into semiconductor products including calculators, minicomputers, and continued work in electronic devices, including making a caesium clock, logic analysers, oscilloscopes etc. Most of this was later spun off into Agilent in the 1990s and 2000s.
"transputers could have been killer GPUs. But alas no."
I spent a few months exploring a B419 - which had a T800 and some VRAM onboard. It was great for driving a nice high-resolution display - but the processor to video memory bandwidth just wasn't sufficent to do full screen animation, if I had been cleverer maybe I could have got some mileage out of VNC style compression to boost the effective bandwidth. The VRAM did offer a mechanism to do row-wise bitwise ops - but that was not really what I needed. The competition at that time was pretty awful too any one remember the i860 ? ... Thought not.
At that point in time with DS links, there was a near future where a Transputer hooked up to VRAM could offer just enough bandwidth to the display to do some high resolution real-time rendering, could have been interesting if it materialized.
There was a famous conference - I forget the details - when Inmos, Intel and I think MIPs were presenting embedded board technology. Intel claimed they could build an embedded board with only 5 support chips. MIPs were next up saying they could build an embedded board with only 3 support chips. Then came Inmos "Why would you need support chips? We just need a power supply!"
... and a 5Mhz crystal ... which was a lot cheaper & widely available compared to the faster ones (in the day). The other nice thing was that you didn't need a common clock or intricate distribution schemes when networking Transputers together - unlike say trying to get a pair of 68K's to run in lock-step. ;)
The MBA is about processes and the costs of these processes. As long as the Marginal costs is below the Average Revenue all is well. Research purchased can fit into this equation but the 9in-house research falls into where on the spread sheet. This was pointed out in the story about the peanuts and the saloon back last century. As noted here the this applies to share value which is the key metric used by my retirement fund in investing strategy. That the product has a flaw is not part of the equation and the liability suits that follow are not to be seen in this fiscal period. The flawed product may be improved at the expense of the MC vs AR ratio but the thought processes engendered by the MBA do not permit this to be easily entertained.
The process in ENGINEERING is that a partial failure is a failure. The product did not perform to all specification: Therefore the product is built with a safety factor (additional cost). This safety factor in theory protects all the foreseeable events but the MBA thought process asks do we need to guard against all events? Why not protect against the most likely and lower the Marginal Cost increasing revenue. A natural adversarial position.
One fly-in-the-ointment that prevents some sort of equilibrium - The Big Institutions. These institutions, mutual, retirement and insurance funds, are required to maximize their return on investment either by participants or government but none have devised a way other than quarterly statement. As a consequence the accounting in the short term beats out Good, New in-house research with its committed engineering.
If you are an engineering company you idealy need a visionary engineer running the company. In house lawers and accountants are functionaries, external lawers and accountants are advisors.
If you are running an accountancy company an accountant would probably be a good option.
If you are running a general trading company then a salesman type may work best.
If you run your company well and provide custoners with good products at a good price and you are not profligate you will probably make a profit which will be adequate to keep your shareholders happy as long as they are not gready.
In my experience, managers, especially 'new' managers, tend to bring in their own people ('yes' men/women) who can only amplify their policies.
Being a small, successful business, then being taken over by a large conglomerate and told "You're doing it all wrong." is a masterstroke in incompetence, misunderstanding and disenfranchisement of employees. You must have to be taught this, but not in an Engineering Degree.
At least we could buy a few beers afterwards ----->
Certainly wasn't taught this during my MBA, what was taught was "business is made up of many skills and experiences. You have your own skill and experience, you'll never be professionally skilled in areas outside your own expertise, what you need to be a good leader is to have sufficiently detailed understanding to talk to those experts and understand what they are saying". Not "how to run a business into the ground", not "how to throw away a good company's heritage/differentiation/unique sales prop", not "count the beans, only the beans matter". Case studies included many famous business failures and mistakes as opportunities to dismantle those, and see what went wrong.
I think people, especially round here conflate MBAs with rapacious corporate finance types. And I know the hedge funds, activist investors, vulture funds, private equity wastrels often recruit new MBAs, but the point is they then train them in their own asset-stripping bad ways.
I'm a manager and I know I'm an idiot. When I get new hires, I always tell them to tell me if they think I'm wrong or they have a better idea.
I also seem to have gained a reputation as someone who can solve problems and get things done. (NOTE: *I* don't get things done, my teams get things done)
I wonder if the two things are related?
The best manager I ever had at a technical job (or any job) was completely non-technical, an ex-military man in fact. He ran the network security group at CalTech's Jet Propulsion Lab (JPL) in the mid-1990s.
He knew how to do three things, and saw them as his job in managing the team: Hire good people, listen to what they tell him, and make sure they have what they need to do their jobs effectively. Whenever that meant running interference or kicking down roadblocks outside or inside the lab, it was always the same response: "Tell me what you need, and I'll make it happen."
Worked for one such as you 20+ years ago. Now retired but if he rang tomorrow and said "Come work for me again." I would - even if it meant international relocation.
He said "I don't understand what you do but I know you do it well so keep doing it, and if you have any problems or need anything let me know.".
Good engineering teams require good managers; one cannot exist without the other.
Isn't this a common pattern? Once there are successful products the cash leeching finance business shenanigans come buy to suck out as much cash as possible. Of course the host will loose his ability to produce the same amount of cash as before since vital investments into the foundations of the success are drained.
All empires fail eventually.
The Roman Empire failed, IBM nearly failed, Boeing is close to failing. In the UK, GEC, Marconi, EMI, British Leyland are all gone.
They were all seemingly too big to fail.
The problem is nearly always the same one - the distance from the top to the bottom is too big. Those at the top have no idea what is going on further down. And those at the top rarely have gotten their hands dirty.
The problem has to be laid at the Company Directors' doors. They took their bonuses and share offerings whilst still enjoying good lunches in executive dinning rooms. They thought that more MBA qualified consultants would fix the problems while they rearranged the deckchairs on the Titanic. And, the big joke is that, when their current business fails, they will still offered well paid part time jobs is other companies. Recently, the Post Office enquiry paraded lots of part-timers which always "had no idea" what was going on lower down. Yet they still got paid - and handsomely too.
Over in Germany Bosch, Siemens, Mercedes-Benz, BMW, Allianz, VW etc are still going strong. Remember, after WWII, we helped the Germans rebuild the VW business and the Americans Honda for the Japanese.
As others have written, there is no secret sauce here; we need good engineers at all levels of the businesses and, particularly at the top.
Yes, we still need bean counters, but bean counters on have one vision which is money. Engineers dream and, sometimes, there are riches to be had.
The decline of British engineering is true, but somewhat over-egged, and most of it can be traced back to intended and unintended outcomes of government policy rather than your presumption that MBAs came in and wrecked everything. So high energy costs and sclerotic planning are severe impediments for any engineering or manufacturing business trying to grow or modernise. Programmes to nationalise, or policies to encourage privately consolidation of entire industries came from government and doomed things like shipbuilding, steel, volume car making, aircraft making etc.
It's new year's eve at the end of 1994. Andy Grove, Intel's boss, is celebrating. It's been a good year, with the new Pentium product doing well.
So he's in a swanky hotel bar, and orders a dram of top-shelf Scotch.
The bartender puts it in front of him and says, "that'll be twenty dollars, sir".
Grove puts a $20 bill on the bar, looks at it for a moment, and says "keep the change".
Ex-Intel here (one of so, so, many... :( )
When I was interviewed to be hired by the company in the late 90s, my interviewer - subsequently to be my boss - asked me what I knew about Intel. Guilelessly like the fresh-out-of-university naïf I was, I said "oh, the FDIV flaw". He twitched visibly and I realized it perhaps wasn't my most tactful comment.
Fortunately they overlooked that, gave me a job, and I had over a quarter-century working for them. A brilliant, exciting, frustrating, incompetent, amazing company, and I don't regret a moment of my time there.
A bit like working for any other large corporate, from the sounds of things! I'm approaching quarter of a century with the current outfit, and can make ALL the same complaints.
Diversion of resources from technical to dealing with administration and accounts is the biggest timesink by far; and given that the workforce split is probably 80/20 against the technical this says a lot about the failures of bureaucracy.
I am *certain* there are folks at Intel with bright ideas. Cut the beancounters out and let the ideas people loose!
What about availability and quality of education?
Student debt is a reminder. This could explain why some EU companies are doing well. Besides who wants to study complex tech and PAY for it.
Then the idiotic visa quotas for US tech companies, while the Mexican border is wide open for the unqualified.
In the IT business, Europe is mainly a failure. Civil Aerospace is excellent with Airbus. Cars are kind of OK. Generally industry is hit hard by insane energy policy. Krupp currently dying. That was once the heart of German industry.
But then, all of the "western world" has a truckload of strategic problems, like not having children and insane immigration practices.
Russia not better, they decided to worsen their similar issues by killing hundreds of thousands of men in a pointless war. A case of Imperial Butthurt.
The number of downvotes on this is really surprising.
The cost of education in the UK and US is astronomical; and the number of good vacancies is poor. In IT sector in particular we've seen practically everything but the very top, and very bottom jobs thrown offshore. It is why I got out of IT. EU and US I am certain are not immune to offshoring problems that business leaders have set us on the road of.
There are a lot of vacancies in engineering and science fields; but they mostly demand experience, and so, the same pool of people are fought over again and again; while never replenishing it with good numbers of entry level posts that meet reasonable cost of living expectations.
The whole thing is deeply frustrating, and I am "lucky" to have avoided the worst that the employment market has had to spit at people of the last 15 years. Corporate leaders echo the difficulties but when do they actually do anything about it...
National government it's even more obvious, with 40 years of shit served on a plate.
I worked at Intel as a software engineer. I had a fun lunch game of bridge with some interesting folks - one was the qa lead for all slot one processors (remember those). Caught an article that Intel had just released the 1.13 Ghz Pentium iii processor.
I mentioned this at lunch and my friend said no we hadn't - well Intel had released it. https://www.theregister.com/2000/08/28/intel_recalls_1_13ghz_pentium/ (way way back machine)
Turns out it did not pass QA and he hadn't signed off on the release. There was an analog bug in the chip that a certain series fo instructions would not result in correct results
Sure enough, Intel recalled all of the chips and didn't get above 1Ghz for a very long time.
Yes, this was a business decision, that overturned the engineer's data that said don't do it
And just like Boeing, some...thing will somehow keep Intel around when they should have long been closed.
Remember the Pentium bug?
Missing the mobile market? (Atom)
Itanium?
IP theft?
Anti-competitive skullduggery?
Custom foundry debacle?
the 10nm production delay because the mfg process was too complicated?
I quit looking after just discovering these.
The problem that recurs is that they install bean-counters at the head of a highly technological company and who have no affinity with technology at all. These people see engineers as quaint, queer and obnoxious, a necessary evil. They only know marketing and playing with numbers to satisfy the stock market.
Intel had a strong candidate to enter the mobile market with StrongARM but instead one of aforementioned clueless CEO's imagined people wanted Windows running on their smartphones and jettisoned the IP to Marvell. Probably propping up the profit for a single quarter and cashing a bonus, but ruining Intel's chances for the long run.
The only good thing that can be said is that Intel saw its mistakes and installed an engineer as its new leader. The mess his predecessors left behind will take a long time to clean up, though, and it's questionable if the "market" will have the patience for him to finish his job.
If you didn't know about these Habana chips, your knowledge of Intel is pretty shallow.
Yes just one more of Intel's acquisitions that went nowhere but it seemed more promising than the larger ones, at the time.
More to the point Intel seemed to be playing three-dimensional chess and acquiring a lot of pieces that might work synergistically.
And you didn't know about it?
Nobody is going to dispute they've found a flaw. It's being patched, at least, for CPUs of certain vintage. We don't know the performance hit of the patch (probably not zero) yet. Comparable to Spectre or Heartbleed I would imagine.
In Intel land we have the 13th/14th gen overvoltage-by-turbo debacle; which is known to be leading to early write offs due to poor technical documentation and integration between Intel & the motherboard OEMs. That's much more significant failure than a microcode patch being needed.
Complicated machines are hard to debug. TOO complicated, it is possible to argue. But what else are you gonna buy if not X86 or ARM...
I have plenty of Intel equipment, but I only buy cast-off server gear that is a) built to run forever and b) completely depreciated to scrap value. My needs are humble, but also happen to be completely fulfilled.
(Another factor: once I gave up Microsoft and Apple, and switched to Linux and open source, I found I had no need for latest & greatest computer hardware. My main server now is an old Datto NAS powered by Intel 4205u running Proxmox, serving several Nginx pages, a NFS server, monitoring and transcoding 4 security cameras, and a whole bunch of other stuff I forgot what it is, purchased scrap value $30 yet running 24/7 at about 35% utilization.)
Intel basically invented the microprocessor with the 4040 and 8080, but since then their track record at being innovative have not been good:
- They planned on the iAPX 432 to be their flagship product, and designed the 8086 mainly as a stop-gap measure. iAPX 432 failed, and 8086 succeeded mainly by being chosen by IBM for their PC. And that platform succeeded mainly because IBM didn't get exclusive rights to the OS, so lots of clones appeared and many PC-makers (such as Commodore) dropped their own product lines to make IBM clones.
- Intel was moderately successful with the 386 and Pentium (which were not really innovative, but clamp-on 32-bit extensions to 8086), but they still rode on the MS-DOS/Windows success.
- Intel then planned on Itanium being their 64-bit platform, but as that failed to gain traction and AMD had success with their own 64-bit variant of the x86 platform, Intel had to make their own versions of the AMD processor.
- Intel, as mentioned earlier, didn't expect the mobile phone market to be so lucrative, and only made a half-hearted effort (the Atom) in making a processor for that market.
- Intel insisted on making complete processor chips instead of selling cores that other companies could add to -- something that ARM did quite successfully.
- Intel has failed to gain real traction on high-end graphics processors, leaving Nvidia and AMD as the main players here.
So, in my book, Intel has since 1980 mainly been successful as a manufacturer of chips and not so much as a designer of same. They have seen the mobile devices market go to ARM, the graphics market go to Nvidia and AMD, lost the Apple processor market to ARM, and are beginning to lose the server market also to ARM. Their great millennium efforts (iAPX 43 and Itanium) failed because Intel forgot to keep things simple.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2025
