Future of Semiconductors and AI Training

A bonanza on the semiconductor industry and hardware scaling to AGI by the end of the decade.

Dylan Patel runs Semianalysis, the leading publication and research firm on AI hardware. Jon Y runs Asianometry, the world’s best YouTube channel on semiconductors and business history.

* What Xi would do if he became scaling pilled

* $ 1T+ in datacenter buildout by end of decade

Watch on YouTube. Listen on Apple Podcasts, Spotify, or any other podcast platform. Read the full transcript here. Follow me on Twitter for updates on future episodes.

Sponsors:

* Jane Street is looking to hire their next generation of leaders. Their deep learning team is looking for FPGA programmers, CUDA programmers, and ML researchers. To learn more about their full time roles, internship, tech podcast, and upcoming Kaggle competition, go here.

* This episode is brought to you by Stripe, financial infrastructure for the internet. Millions of companies from Anthropic to Amazon use Stripe to accept payments, automate financial processes and grow their revenue.

If you’re interested in advertising on the podcast, check out this page.

Timestamps

00:00:00 – Xi's path to AGI

00:04:20 – Liang Mong Song

00:08:25 – How semiconductors get better

00:11:16 – China can centralize compute

00:18:50 – Export controls & sanctions

00:32:51 – Huawei's intense culture

00:38:51 – Why the semiconductor industry is so stratified

00:40:58 – N2 should not exist

00:45:53 – Taiwan invasion hypothetical

00:49:21 – Mind-boggling complexity of semiconductors

00:59:13 – Chip architecture design

01:04:36 – Architectures lead to different AI models? China vs. US

01:10:12 – Being head of compute at an AI lab

01:16:24 – Scaling costs and power demand

01:37:05 – Are we financing an AI bubble?

01:50:20 – Starting Asianometry and SemiAnalysis

02:06:10 – Opportunities in the semiconductor stack