ThrowawayLegalNL

joined 11 months ago
[โ€“] [email protected] 1 points 11 months ago

I half-agree with this. China has definitely been pushing for a domestic chip industry, but the Huawei debacle, and last year's October surprise lit a fire under their feet. Both in terms of state funding, but especially because it forced the private sector to make changes.

We see China as this top-down governed behemoth, but this does not capture the whole picture. Most private companies operate in a relatively autonomous fashion. That is why until recently, most tech companies (Tencent, Baidu, ByteDance, even Huawei) were not interested in domestic chips: because they were uncompetitive. So they bought Nvidia, Qualcomm, and TSMC chips. Now, all their funds are going to domestic manufacturers, which is bad for the tech companies, but good for China's goal of chip autonomy.

So my main argument is that China was indeed already doing a ton to decouple from the West before (and during the initial phase of) the sanctions, but the private sector has accelerated this shift out of necessity.

This does not mean that sanctions were a bad idea from the US perspective. Absent those, something like M60 Pro may have only been developed by 2026, but China would still be transitioning towards indigenous chip production -- while also reaping the benefits (AI race, increased global profits bein funneled into R&D) of access to foreign tech.

[โ€“] [email protected] 1 points 11 months ago (4 children)

At the very least, it is causing some short- and medium-term trouble in China's AI advancements. China has to build its own industries and architectures as they're cut off from high-end processes in the West and SK/Taiwan. Companies like Alibaba and ByteDance used to buy Nvidia because domestic alternatives were uncompetitive. Now they're forced to buy domestic, which is very good for China's aspirations of Chip autonomy, but problematic in the short term.

The upside (if your goal is to hobble China) can only be evaluated in the future, based on how much China can catch up using indigenous processes. My impression is that they're doing quite a bit better than expected, but obstacles (EUV especially) remain.