A particularly surprising discovery is that CXMT is constructing its cleanrooms—where the actual silicon is produced—in just around 12 months, whereas the rest of the DRAM manufacturing industry typically takes 21-24 months, nearly a full year longer.
Finally, the report notes that the Chinese government is encouraging CXMT to transfer its DRAM technology IP to other Chinese DRAM manufacturers such as JHICC, Swaysure, and YTMC’s subsidiary XMC. This initiative aims to alleviate memory shortages domestically before CAMT-designed DRAM enters other markets, such as the EU and possibly even the U.S.
This shit is so over. This gang ain’t gonna collude on price with Micron. They’re in for eating Micron. First its market share, then its IP by buying it for pennies on the dollar.
Especially pay attention to the CCP’s directive to share IP among RAM manufacturers. They appear to be looking at this problem as a supply shortage of an essential component for the rest of the economy, which it is. But that also means they aren’t playing for profit.
aren’t playing for profit
CXMT is probably the most profitable tech company in China right now.
I meant the CCP isn’t, not CXMT.
But the country is not playing for profit of this single company, which would probably be the case in USA (as the company would have lobbying power yo secure that)
Micron doesn’t seem interested in competing for consumer memory anyway, since they exited the consumer memory market. Latent demand exists for memory at not-outrageous prices so good for CXMT for filling that niche.
Do you think they’ll stop expanding after filling the consumer memory market though?
I don’t think so, domestic electronics manufacturers have an enormous need for memory that is performant but doesn’t necessarily need the top-of-the-line, and probably China’s government is pushing to reduce the need for imported memory components from Korea or Taiwan as well, so demand for production is there coming from both the private and public sector.
That’s what I’m thinking too. They’re gonna go as far and as wide as they can go.
Micron DDR chips are still being added to many consumer devices. They just would down the side the sold SSDs and similar direct to consumers.
The companies are still focusing on larger customers that are willing to pay more, one has to wonder how they will fare when those businesses aren’t willing to pay as much and the companies they told to take it or leave it have options from China.
Having actually done this, the clean room part is super easy. Getting all the equipment in and running is extremely hard. Finding out all the things you messed up along the way and getting actual yields is extremely hard and also time consuming. So I’d like to know where along that process they are doing the timeline comparison.
I mean… Also assuming this is all real, and the rushed schedule doesn’t produce low quality botched RAM sticks that die in six months.
When it comes to products from China, they can be incredible; unprecedented quality for so much cheaper! Or they can be exactly what everyone says they are and end up in a landfill after a short amount of time.
It’s quite the coin flip.
“assuming all current projects are completed on schedule”
From the article. This is the single largest fallacy one can engage in in engineering.
“CXMT is projected to produce about 350,000 DRAM wafers per month by the end of 2026, assuming all current projects are completed on schedule.”
This is a small blurb on a random tech site, it’s just giving a simple ‘if all goes well’ projection.
I’m sure they’re not setting the entire undertaking up so that if one thing ends up slightly delayed everyone’s heads will explode. It seems likely that they also have projections that take varying delays with varying parts and numbers of parts of the project into account, but that someone writing a small blurb on a tech site isn’t really privy to the entirety of their forecasting.
I don’t think they need you to rush in and save them from their folly.
Why would you assume any of that? If there is one constant in large-scale engineering projects, it is project managers over-promising on what their teams can deliver. In this case, they are promising a miracle salvation from the global DRAM shortage, so a measure of scepticism is, I believe, warranted.
Everything, everywhere, doesn’t always happen as it does in your frame of reference and experience. I find it unlikely that you’re genuinely an expert in engineering projects of this sort and specifically in exactly how they’re undertaken in China with large scale government support.
I am not an expert in engineering projects in China, which is why I rely on 3rd party reporting to give me an idea of what is going on in that space. And when China makes what seems to me to be a preposterous claim, qualified by “if everything goes according to schedule,” I am suspicious. Not because I particularly distrust Chinese engineers, but because the project managers in my own country are prone to the same tendency to overstate scope, and understate budget and schedule.
Competition is good. Hopefully an end to memory manufacturer price collusion is near. I’m not holding my breath for it though.
Watch international* ram be banned like international* EVs
*Chinese
Didn’t this admin already warn Apple?
It’s better to use their weak labour standards to exploit their workers for low wages, rather than try to compete with those conditions directly.
