Can’t remember the full details of the deal, but I seem to recall a story about how Apple approached Intel to manufacture a low-powered processor for mobile (for the first iPhone). At the time, Intel didn’t see money in mobile processors and passed on the deal. Additionally, for years, Apple asked for more powerful chips for the MacBooks. At the time, the iPads were surpassing MacBooks in speed on some tasks. Finally, Apple decided that since they were already designing their own silicon for iPhones and iPads, they might as well just do the same for the MacBooks as well since Intel couldn’t keep up.
Again, this is largely from memory. I can’t remember the source, so take it with a grain of salt.
I’ve only watched this from a great distance but what I saw was: Intel didn’t actually manufacture the chips. That was all TSMC. So Intel’s main thing was chip design. And their designs were all about making the transistors smaller. Around 3nm they started running into physical limits. Competitors started out-innovating them with things like GPU deigns and ARM based chips. End of story. They had their time. They ran x86 into the ground and they are fucking done. They would have had to do 5 or 6 things differently to stay on top, and they did none of those.
They always had their own fabs. Utilising TSMC for their GPU’s was a recent thing. For all the mistakes they have done, their GPU efforts are actually noteworthy but you don’t even have to compare them to ARM or other GPU manufacturers, just look at AMD, they’ve been killing it.
Intel fell behind on chip manufacturing while the CEO came from that department.
Allegedly because their strategy was too ambitious at the time, or at least that was the official excuse at the time.
So your summary is not entirely fair.
Well, when I was learning about economics being 8 or 9 year old, it seemed for me how it should be.
A person or a group knowledgeable in some area find a bottleneck, some problem to solve, start a company, it grows, it becomes big. Then the next generation is what they pick for leadership, and picking people is always worse than the evolutionary mechanism of a company finding some bottleneck to be widened being gunshot faster than the rest. Then they pick their replacement. And so on. Eventually it dies, but since technologies are patented, they do not become actual secrets, only commercial secrets, and by the time a company dies the patents expire, so everybody can replace it for the humanity.
The niche that company discovered thus becomes competitive.
In our world, if patents would expire as fast as they did initially by design, these big companies would already be dead.
But they’ve bent the rules to make patents virtually eternal and thus big zombie companies are strangling the humanity.
The system wasn’t bad, but eventually power changed it.
You are missing economies of scale. In most industries these create a significant barrier to entry. The patent may expire but the equipment is still expensive.
Especially abominations like patenting an ISA. It’s clear from the very beginning that an ISA is not an invention moving humanity forward, it’s an interface. A language.
As of gigantic companies of today not finding replacement when they die - we would have the whole spectrum if not for IP and patent laws as they exist. For some uses MCs of 80s are sufficient. For some a desktop PC of 1993. For some a desktop PC of 1999.
I dunno why I’m writing these things, Marcus Aurelius has written many wise things, one of them is the advice not to think about things out of your control.
Sounds similar to what happened to Boeing. Once ran by engineers now ran by people suckling the teat of board members. Quality goes down, profits go up for these assholes.
Just watched a video on the failure of windows phone, they went from 34% market share ( world top 1) to 1.4% in 5 years. Then they recover a little bit to 3%, just to drop to 0.4% 5 year later and then completely dead 2 years after.
Never at any point in time did the Windows phone reach 34% market share or anywhere near #1. I’m not even sure Windows phone had a bigger share than BlackBerry at the time.
Their peak market share was 3.4%, not 34%. It failed because virtually nobody bought them.
The only piece of Microsoft tech that I actually loved, so sad it flopped. I had two Windows phones, beautiful devices. Gorgeous screens, great design, the Windows 8 tiles unironically were fantastic on mobile.
Everything was butter smooth, I never had them crash or freeze up. Zeiss cameras, they took great pictures.
But there were almost no apps for them. It was basically the Microsoft mobile office suite, and a few random ports like Evernote. Nobody bought them because there was zero ecosystem for them.
That’s not “Windows phone” that’s “Windows mobile”, the precursor to Windows Phone, which didn’t release until 2010.
Shifting to Windows Mobile now, in 2006, Windows Mobile 6 had only about 10% market share, behind both Palm OS and Symbian, the latter of which held a whopping 60%. I looked further back in time and I do see that Windows Mobile had a 34% market share in 2001, however it was again dwarfed by PalmOS. It’s also worth it to note that that 34% wasn’t comprised mainly of cellphones, but rather barcode scanning guns in warehouses and logistics, because you could make custom applications for them with relative ease. There are still warehouses today that use those old windows mobile scanner guns.
Probably bureaucracy. Also an inability to pivot even when things make no sense. Everything is a giant freight train that has very little ability to change direction or stop.
Oh and of course a healthy taste of not being transparent or honest.
This happens easily for big successful organisations. Over decades a strong culture aligned with how they succeed forms. Once the market changes requiring a culture change, a seemingly invincible company suddenly stumbles. They simply can’t respond even if they what they should change.
Ex. Rolls Royce CEO stated this phenomenon well: culture eats strategy for breakfast.
Anything they go after today is 18-24 months out. Chasing after AI would be pretty risky. Desktops and laptops are moving to ARM and RISC-V. Their best bet is to go after whatever enterprise data centers will need a couple of years from now.
If I were laying bets, it would be to go after power and heat efficiency. Like, hard. Take their time out in the wilderness, then come back with chips that save the planet from climate collapse.
Oh they won’t die. The question is will they recover to their old market position, will they downsize and be second fiddle to AMD but remain generally profitable, or will they have a slow managed decline like IBM?
I think IBM was different because its lunch was eaten almost entirely by other American companies (chiefly Microsoft). That probably wouldn’t be the case if Intel were allowed to declined in a similar manner.
hardware and software have taken turns in long waves for 50 years. like for self driving cars right now, the hardware is ready but tue software is catching up. intel hasn’t led the bitcoin/ai waves, and microsoft is no longer married to intel, and gaming and mobile phones aren’t intel either. they are late to RISC/ARM, etc. they are too big to survive on niche and they are missing lots of major waves.
The same way boeing executives are to blame. They did not order employees to do sub-par cpus, but they did not care about the quality of what was produced either. Good hardware (and software) its always the product of a process that involves QA, HR, Operations, R&D and many other departments. These departments fall under the supervision/control of the executive suite.
Well run cost center departments don’t boost quarterly results, ergo they are deprioritized.
They are looking out for themselves rather then the company, because of the incentives in place.
We are living in weird times where stock price doesn’t really correspond with company health, so their actions reflect against that metric against all others.
I think it was a mindset shift. Right now short term stock value is more important because the shareholder profile also shifted from someone that holded the stock to someone who wants just to turn a quick profit.
There’s a term in tech called “empire building” where middle management looks for promotion up the chain towards directorships or VP roles. If, for example, you have a CEO that’s nuts for AI all you’ll want to do to get on their good side is to build a team around AI for some random service you already have (e.g. AI in Google Search) and you’ll get a ton of funding and HC. Suddenly you run a huge division and get a fancy important title because you can shit some metrics about how well you’re performing while customers say “wait, search is shit now”. That’s the search team’s problem, your AI stuff performs great!
It’s everywhere in big tech, and it’s why so many big tech companies seemingly work extremely hard and have nothing to show for it.
The IC’s at the bottom of the ladder are just minding their own business, trying to do the best work that they can, while the leader of the empire sets ridiculous timelines and goals because they’re trying to cement a legacy, rather than build the right thing. Naturally, the product flops, the director gets moved to a new division to protect them, and the IC’s are laid off - with the CEO saying that they didn’t meet expectations or cost too much.