translation: AI bubble pop scheduled for 2028?
Not following through on expansion plans indicates that they’re expecting the bubble to burst
They’re most definitely hedging their bets.
If there’s a shortage and they don’t expand, they can charge more per unit. If they expand capacity and there’s no shortage, they’ll either have a bunch of unused capacity or memory will get cheaper than they’d like.
I expect them to end this year. There is no demand by normal customers, only AI focused enterprises.
The whole plan is to make personal computing unaffordable so we have to rent cloud services to do anything.
They’ll find a way to keep that consumer demand at zero.
They want a chunk of your income to gatekeep the products they made unaffordable to you.
Billionaires shouldn’t exist.
They’ll have to pry my DDR5 out of my cold dead motherboard before that happens. Tech has stagnated so much that my last rig lasted 12 years from 2012 to 2024. This current rig will outlast this entire administration.
I do feel bad for all of you who want or need a new build now. That’s where it hurts the most.
Either that or, optimistically, memory manufacturers see a bubble about to burst, and they don’t want to be stuck holding excessive manufacturing equipment.
Frequently the attitude towards equipment is that it can be sold once they’re done with it. Particularly specialized equipment might sit a little longer but that’s built in to the calculation.
There’s a bottleneck in hiring skilled workers to run that equipment, and time to get that equipment up and running. It can take months from when the equipment is in the building before it’s making parts, up to a year for very sensitive equipment. And troubleshooting it to get it working at capacity takes even longer.
Which is great, if you need time to find people to run it, but usually those skilled workers already have jobs, and there are only so many available in any given area. Even if you have them available you’re stuck paying them for months before you’re ready to put them to work.
It’s high risk and it takes a long time to pay off. A miscalculation is the end of the CEOs career and may bankrupt the company. The first to pull it off though will be in a great spot, assuming it’s not a bubble and it doesn’t burst. Manufacturers are going to want some solid guarantees before taking the plunge. Optimally they want someone else to pay for it. They may actually be negotiating with these companies now about funding their expansion.
Maybe. I mean at least a major part of it is that AI pays a lot more money than the consumer does, and even if the consumer market pulls back, they are banking on the raised prices to stay around even if they can increase capacity.
I think what scares them is that CXMT is rapidly catching up to the state of the art. If they dick around for too long, they run the real risk that China and CXMT will do what China does and sweep the market with really cheap memory they can’t compete with.
These companies still care about non-ai servers, and that’s a big part of the market that could be obliterate pretty quickly.
Given CXMT doesn’t just up their prices after gaining market share too. Also China can just ban the export, or tariff it.
Given CXMT doesn’t just up their prices after gaining market share too.
They might eventually, but certainly not immediately. In the process they are going to force the other big players to lower their prices to compete.
Also China can just ban the export, or tariff it.
I don’t think China has ever banned an export. It’s pretty rare to tariff an export, basically only happens when it’s a limited good that the government wants to ensure a local supply of. I think the only country I’ve heard of doing that is Greenland due to some british wankery.
CXMT (or whoever is sticking the memory onto PCBs) already have increased the prices.
Just like AMD, there’s no reason to actually compete on price when you can offer close enough for 10% less.
Technically speaking, Tariffs only exist on imports. Export taxes are an entirely separate category of tax. You are correct though, the logical thing for China to do is to run at a state-subsidized loss with CXMT until everyone else has left the market before raising prices. You know, basically pulling a Walmart with state help.
That isn’t the plan. Person computers are still available, but ram production didn’t match demand.
Also billionaires don’t have a billion dollars so they don’t effect you. Their wealth is stocks that only home value because you keep buying them. If the bubble pops those “billions” disappear.
Most companies don’t make much money selling to “normal” customers. Enterprise sales are almost always the bulk of the revenue.
B2B sales are much less of a headache than B2C. You can make fewer, larger sales, and you don’t have to deal with The Public.
Why would that mean it ends this year?
New consoles and maybe new steam deck. Maybe raspberry pi and similar SBC hardware won’t be stupid expensive anymore
