2023-03-08 19:49:24
2023-03-07 08:37:31
2023-03-07 08:37:14
5785391
The market for PCs has stalled so much that I'm still using a 2015 Dell XPS as my main driver - for comparison, try to imagine using a 1996 Pentium machine as your main driver in 2004.
A new laptop with similar specs to mine would still cost at least $800 today - again, try to imagine a world where a Pentium One lost almost no market value eight years later.
8GB of RAM is still the norm, and today's CPUs have just tiny improvements over those produced 8 years ago.
In the past eight years, technological innovation in the industry has literally plateaued, with everyone going after easier-to-monetize mobile users rather than thinking of those who want a big screen.
And it definitely didn't help that the manufacturer who held a total monopoly on the market (Intel) has been managed in an embarrassing way for the past few years, has missed all the trains, has actively lobbied to keep its dominant position while delivering literally zero added value, and it's now in a state where, despite billions in generous government subsidies, despite a position of total monopoly consolidated over seven decades, despite having hired the most brilliant mind in electronics for decades, its market share is challenged by competitors such as AMD, ARM, RISC-V and Nvidia.
Things have gotten so depressing for this industry that the only way left to boost product sales is to sunset Windows 10. In other words, to force customers to swallow planned obsolescence in order to force them to buy a new computer.
Whoever says that capitalism fosters innovation and competition has clearly no clue of how the world works.
https://www.theregister.com/2023/03/07/pc_sales_worse_idc/
A new laptop with similar specs to mine would still cost at least $800 today - again, try to imagine a world where a Pentium One lost almost no market value eight years later.
8GB of RAM is still the norm, and today's CPUs have just tiny improvements over those produced 8 years ago.
In the past eight years, technological innovation in the industry has literally plateaued, with everyone going after easier-to-monetize mobile users rather than thinking of those who want a big screen.
And it definitely didn't help that the manufacturer who held a total monopoly on the market (Intel) has been managed in an embarrassing way for the past few years, has missed all the trains, has actively lobbied to keep its dominant position while delivering literally zero added value, and it's now in a state where, despite billions in generous government subsidies, despite a position of total monopoly consolidated over seven decades, despite having hired the most brilliant mind in electronics for decades, its market share is challenged by competitors such as AMD, ARM, RISC-V and Nvidia.
Things have gotten so depressing for this industry that the only way left to boost product sales is to sunset Windows 10. In other words, to force customers to swallow planned obsolescence in order to force them to buy a new computer.
Whoever says that capitalism fosters innovation and competition has clearly no clue of how the world works.
https://www.theregister.com/2023/03/07/pc_sales_worse_idc/
IDC gets even more pessimistic about PC sales
Retirement of Windows 10 should eventually drive recovery in 2024 and 2025Laura Dobberstein (The Register)
Niclas Hedhman
in reply to Fabio Manganiello • • •And yet, your final statement is contradicting your 5th paragraph, where you say competition is challenging and dethroning Intel. How? By innovation.
And if we didn't have a State to be lobbied for handouts, tax-breaks, subsidies and regulation, it would have been harder for the giant corps to hang on to their power.
Fabio Manganiello
in reply to Niclas Hedhman • • •In a purely capitalist market, that happens when a business has reached such a size and market share that the entry barriers to compete with it are just too high for anybody to actually be able to compete with it, and the regulators did nothing to prevent that business from growing so fat.
On the public side, it happens when administrators pick winners instead of will.
Funding one specific company that has "great strategic importance" and is "too big to fail", like Biden did with Intel, is fundamentally wrong. Just like Macron is wrong when he talks about "fostering European champions" that can compete with American giants - i.e. pick those who have the blessing to grow huge instead of coming up with a technological vision and rewarding whoever helps implementing it.
It would have been much fairer to announce a plan (e.g. "we want to build a factory on American soil that can produce chips at scale with a 3 nm litogr... show more
In a purely capitalist market, that happens when a business has reached such a size and market share that the entry barriers to compete with it are just too high for anybody to actually be able to compete with it, and the regulators did nothing to prevent that business from growing so fat.
On the public side, it happens when administrators pick winners instead of will.
Funding one specific company that has "great strategic importance" and is "too big to fail", like Biden did with Intel, is fundamentally wrong. Just like Macron is wrong when he talks about "fostering European champions" that can compete with American giants - i.e. pick those who have the blessing to grow huge instead of coming up with a technological vision and rewarding whoever helps implementing it.
It would have been much fairer to announce a plan (e.g. "we want to build a factory on American soil that can produce chips at scale with a 3 nm litography process") and then grant the funding to whoever presented the most sensible plan at a public tender, rather than just handing out money to a company that has achieved NOTHING in the past decade, just because they're still the biggest. This goes against any idea of meritocracy.
All in all, I'm happy that my 8-year-old laptop is still doing a good job today, but at least I'd expect to have some alternatives on the market today that are much better and would at least make me consider an upgrade.
Instead, most of the laptops on the market today have specs that aren't much different from my laptop, and that's the problem: why would anybody spend $1000 to buy something that is basically a copycat of what they have already?
It eventually becomes a self-fullfilling prophecy: people don't buy new laptops because the old ones are still doing fine and the new ones don't offer many exciting features to justify the investment. As people are less incentivised to buy a new product, the revenue stream for these companies dries up as well and they have less money and resources to invest in innovation. As they invest less in innovation, their products become more mediocre and people have even less incentives to upgrade. Until the only way left to push people to buy a new product is through the largely immoral planned obsolescence pattern.
≠
in reply to Fabio Manganiello • • •The technology has peaked or pleateau'd; the innovation has gone elsewhere.
Specifically to mobile devices which are a much bigger market.
Fabio Manganiello
in reply to ≠ • • •That includes both the "free market" that has let companies like Intel grasp nearly 100% of the market with no competition and no external intervention (ending up in a situation where they can stop innovating for a whole decade while losing no market share), as well as acts that diverted a lot of public money to Intel just because "they're the biggest", even though they have done nothing to deserve such a boost and they have no plan outside of "we'll build a big factory and hire a lot of Americans".
I don't think that the technology's plateau is inevitable. Sure, going smaller than 5 nm is a physical challenge, and Moore's law has already been challenged a lot, but there are companies out there that are managing to do it. We have the means to hit a sweet spot where old hardware can still work fine for many years, but people also have enough incentives to update to newer hardware. The pro... show more
That includes both the "free market" that has let companies like Intel grasp nearly 100% of the market with no competition and no external intervention (ending up in a situation where they can stop innovating for a whole decade while losing no market share), as well as acts that diverted a lot of public money to Intel just because "they're the biggest", even though they have done nothing to deserve such a boost and they have no plan outside of "we'll build a big factory and hire a lot of Americans".
I don't think that the technology's plateau is inevitable. Sure, going smaller than 5 nm is a physical challenge, and Moore's law has already been challenged a lot, but there are companies out there that are managing to do it. We have the means to hit a sweet spot where old hardware can still work fine for many years, but people also have enough incentives to update to newer hardware. The problem is that other companies don't have the same market share and lobbying power as Intel's, and if they decide to hit the brakes then the whole industry hits the brakes.
So we're in a state where innovating has become more risky and expensive, those who own almost the whole market share have no incentives for taking on more risk (they get public funding even when they do nothing, and getting computer producers to swap Intel for other architectures is a process that takes years), and they still have some cards up their sleeves to boost profit without innovating (like making the software more bloated or going down the route of the dear old planned obsolescence).
Sure, it's not capitalism alone that has taken us into this state, but it's the sum of all the forces that have consolidated an absolute monopoly with no incentives to keep innovating. The idea of a free market without any external intervention definitely helped Intel become a winner-takes-it-all monopoly that makes even the Big Oil and lightbulbs cartels of the first half of the past century pale in comparison.
Sir Viber Ph.D, M.D.
in reply to Fabio Manganiello • • •Try running 3 screens at 1440 with that little Prius of a laptop.
Fabio Manganiello
in reply to Sir Viber Ph.D, M.D. • • •≠
in reply to Fabio Manganiello • • •In particular, nobody really likes Windows 11 and Office 365, except for very simple tasks.
The technology plateau is more complex. I think the bottleneck here is being able to use multicore processors efficiently. We are getting marginal speed gains instead.
teknomunk
in reply to ≠ • • •Fabio Manganiello
in reply to teknomunk • • •teknomunk
in reply to Fabio Manganiello • • •If everything was being run in a single core, I agree that would be useless.
However, we arenct stuck with deciding between single-core single-threaded and the fine granularity multithreaded applications that people fantisize about but that are incrdibly difficult to actually implement. Multi-core is perfectly fine with running multiple applications at the same time whenre none share memory and only communicate thru pipes, sockets or the file system and this style of programming has the benefit of being far easier to write, debug and maintain.
≠
in reply to Fabio Manganiello • • •dew_the_dew :verified:
in reply to ≠ • • •Fabio Manganiello
in reply to dew_the_dew :verified: • • •I believe in demand-induced progress: give people better/more powerful tools, and they'll come up with something you couldn't even imagine earlier.
Crowley :verifiedbi:
in reply to Fabio Manganiello • • •