241–250 of 526 posts
Earlier quoted context omitted.
I guess we are the redundancy to taiwan soon...
One of the redundancies, and not a very high end one
Earlier quoted context omitted.
The fab is for chips that are something along the lines of 4 generations old, and that's ignoring that they won't start manufacturing until 2027... This has absolutely nothing to do with AI or LLMs. Generally these are automation chips for machinery or vehicles. "Microsoft made him walk that one back real quick" This sounds like a weird fan fiction. OpenAI openly negotiates with EU governments. No one walked back any…
The technology is roughly 10 years behind, but in the scenario that the replacement capability becomes important is something like Trump fumbles the great power competition and US and China have a nice nuclear moment. And frankly making do with Haswell era CPUs is lamentable but as far as nuclear exchanges go not that bad.
Earlier quoted context omitted.
The fab is for chips that are something along the lines of 4 generations old, and that's ignoring that they won't start manufacturing until 2027... This has absolutely nothing to do with AI or LLMs. Generally these are automation chips for machinery or vehicles. "Microsoft made him walk that one back real quick" This sounds like a weird fan fiction. OpenAI openly negotiates with EU governments. No one walked back any…
The technology is roughly 10 years behind, but in the scenario that the replacement capability becomes important is something like Trump fumbles the great power competition and US and China have a nice nuclear moment. And frankly making do with Haswell era CPUs is lamentable but as far as nuclear exchanges go not that bad.
For those unfamiliar with the region (Saxony, Dresden): A registered association ( e.V. in German) bluntly named "Silicon Saxony e.V." (est. in 2000) with over 450 members headquartered in Dresden pretty much sums up the infrastructure and ambition.[0][1] As pointed out the process TSMC’s 28/22 nanometer planar CMOS and 16/12 nanometer FinFET process technology (fabrication will begin in 2027) is not "cutting edge" b…
My take, if you want to have a bleeding edge semiconductor industry in your country, first start by trying to build a boring, vanilla, outdated semiconductor industry. If you can do that, then try incrementally making it competitive with global markets. (Yes, this requires tariffs initially to bootstrap the process, otherwise nobody will by your shitty overpriced semiconductors.)
[0] https://astralcodexten.substack.com/p/book-review-how-asia-w...
Earlier quoted context omitted.
> And as soon as China-Taiwan chapter is over That's a very strong prerequisite though… It may not be over for several decades to come.
After what happened to hong kong, its hard to believe that.
- Hong Kong belonged to China already, it just had a special status that was supposed to expire at some point, Xi “just” went ahead of schedule but there wasn't anything to stop him from doing so (no independent HK government, no army, etc)
- Taiwan is a complete state, with a government and an army, a navy and an air force equipped with Western and mordern home-grown equipment, cruise missiles and anti-ship missiles and so on. Without a Western direct armed intervention, Taiwan is almost always expected to fall after a few weeks by Western military analysts (though the same analysts also expected Ukraine to be unable to match Russia on conventional battles), but even the current predictions on Taiwan suggests that China will pay a signifiy cost in terms of casualties and and material even if they succeeded eventually. And again the success assumes no military actions from the West. Of course Xi may take his chance anyway, but the situation would still be incomparable to HK.
Earlier quoted context omitted.
> Bear in mind that this could be a "thin edge of the wedge" situation. Where they get this "older" node stuff up and running, and if it turns out to be going well then they can think about newer process/node sizes. I think that would be more credible if Global Foundries didn't already have "Fab 1" up and running that has a mature 22nm process.
But Global Foundries doesn't have the ability to do newer/smaller node sizes, so they can't "level up" at a future time. That is, unless Global Foundries makes some massive investments. Which they've not demonstrated any willingness to do, nor competence at doing. This deal with TSMC at least seems like it provides a way forward with people that know how to execute. One that the Global Foundries option can't credibly…
But TSMC isn't willing to make those investments in Europe either. It's their goose that lays golden eggs, and keeping it all in Taiwan makes a lot more sense from a geopolitical perspective. The only reason they are now building this fab in Germany is because it is ancient technology to them and they are basically getting a free fab out of it.
Just want to share these relevant videos from Asianometry: How Semiconductors Ruined East Germany https://youtu.be/cxrkC-pMH_s Why Europe Lost Semiconductors https://youtu.be/5ZdmS-EAbHo Europe’s Lost Decades in Semiconductors https://youtu.be/PpSdytrNEBg It's great to get chip manufacturing back to Germany after all the struggles, even if it's not cutting edge.
I watched a few of their videos, but I felt like it was an anti-china propaganda outlet. I'm def not a fan of the CCP, so I loved watching the videos, but I felt like I was getting brainwashed. Does anyone know how valid/factual their videos are?
Earlier quoted context omitted.
Just weird to me that they build in Germany. Germany basically pressed "delete" on (clean) energy for the coming decades and has very little capacity to seriously expand their energy conversion.
Germany has build more PV capacity in the last three years than the total nuclear capacity it ever build, and the numbers are increasing thanks to regulatory cutbacks by the current government. Meanwhile new nuclear reactors cost 4 times as much as renewables and take decades to build. So stop spreading FUD.
For those unfamiliar with the region (Saxony, Dresden): A registered association ( e.V. in German) bluntly named "Silicon Saxony e.V." (est. in 2000) with over 450 members headquartered in Dresden pretty much sums up the infrastructure and ambition.[0][1] As pointed out the process TSMC’s 28/22 nanometer planar CMOS and 16/12 nanometer FinFET process technology (fabrication will begin in 2027) is not "cutting edge" b…
In How Asia Works[0], they contrast how Malaysia tried to modernize by jumping straight to high-tech steel factoris with how South Korea modernized by starting small – basically by making shitty cars – then scaled their way up. Turns out the latter simply works much better. My take, if you want to have a bleeding edge semiconductor industry in your country, first start by trying to build a boring, vanilla, outdated s…