We think there will be a few more legs of semis consolidation. So we compiled a list of 5 semis companies who we think will survive:
1. ADI & TI
2. QCOM
3. NVDA
4. Some chip company from China
5. The smoldering remains of Intel, probably owned by others.
We are all familiar with software and Internet companies building their own chips. Less appreciated is the extent to which semis companies are all gradually morphing into software providers as well.
RISC V is going to encounter a lot of software incompatibility problems, but that probably won’t matter, and definitely won’t stop it.
The market for base station and RAN silicon is shaping up to one of the most interesting to watch in the new year. And intel seems to have forgotten its long, sorry history in the space, making for some great viewing.
TSMC has had decades of government subsidies and currency suppression working in it favor, but they have used that to build an incredibly deep reservoir of human capital which no one else can readily replicate.
The US has a number of ways to encourage allies to support its semis restrictions on China. These range the “stick” of enforcement to the “carrot” of waivers and targeted expansion of the restrictions.
It seems very likely that most AI Training will be run on Nvidia GPUs. The software environments for training are too numerous and changing rapidly, which favors the soft, warm familiarity of CUDA over speculative gains offered by new entrants.
Lots of people think chip companies should build “services” businesses to capture more value. The list of companies that have accomplished this is very close to zero.
The latest US semis restrictions seem to be a dusted off version of 1970’s Cold War measures, but the situation is so different we have a hard time seeing them proving effective and highly likely to spur strong counter reactions in China.
Analog companies have many tricks they use to build almost custom semis, and it seems likely that the digital chip makers will start to emulate some of these practices more often.