Dropping a podcast with @pequityresearch Mr. P has been doing some great work breaking down the AI buildout and following where hyperscaler capex actually goes. In this episode I wanted to walk through the full stack with logic, memory, power, and networking. We get into why memory could become the largest line item in the AI bill, what old GPU rental prices tell us about compute demand, and where value is going to accrue as the physical constraints get harder to solve. At the center of the conversation is his view that AI spending cannot grow forever. Long-term contracts might change the shape of the next memory downturn, but they don’t eliminate the cycle. And signing a 10-year contract does not mean anyone can actually see 10 years of demand. We also get into why he’s excited about optics, where NAND and HBF fit as agents use more memory, and his views on Chinese open source and the future of US model development. We discuss: - Why he thinks 10-year demand visibility is bullshit - Memory’s growing share of hyperscaler capex, and why estimates vary so much - Why older GPUs are still renting and what that says about compute demand - How long-term agreements, pricing floors, and prepayments actually work - Power as a bottleneck, and why identifying a constraint isn’t the same as finding an investment - Copper vs optics, and where networking value accrues - Agents, NAND, and where HBF fits in the memory hierarchy - CXMT, Chinese open source, and the risks of slowing frontier model development Timestamps: 0:00 – Why AI Spending Can’t Grow Forever 1:13 – P Equity Research’s Background 5:48 – Where Hyperscaler Capex Actually Goes 8:00 – Is Compute Still Tight? 12:08 – Memory’s Share of the AI Bill 17:20 – Why the Memory Cycle Isn’t Dead 18:37 – Inside a Long-Term Agreement 28:30 – What Happens If Customers Cancel? 32:17 – The Rising Cost of the AI Buildout 33:43 – Copper vs Optics 38:49 – Power and Gas Turbines 39:29 – US Models and Chinese Open Source 42:36 – Agents and NAND 47:00 – Where HBF Fits 52:23 – What P Is Most Excited About 56:27 – CXMT and China’s Memory Industry 1:04:00 – Closing Thoughts Enjoy!

$MU preview by Wellsf.. earnings coming up. Friends returning from Asia tell me memory sold out till 2031. For whatever it’s worth (told by memory bulls). Do your own DD. NFA.

Massive problem. As my friend Big Ed @edzitron proved most GPU sales are GPUs that are sitting in warehouses. Goes without saying we are power constrained and anyone who can bring power online (or reduce power consumption) is king. Any watt being used for anything other than AI right now is misallocated !!!

Had an exciting podcast with @LoganJastremski to discuss hyperscaler capex, memory, etc. Highly recommend listening to our conversation and giving out your thoughts. It is also up on YouTube.

SemiAnalysis's Jordan Nanos says CoreWeave is so backed up a customer with $100 million to spend was told to wait until May, so Nebius has been capitalizing "They're serving this middle tier of the market. CoreWeave has set the standard. I mean, they are the best in terms of technology, this reliability stuff I'm talking about. They just have so much data for monitoring." "I mean, these guys deploy 10,000 GPUs a week at the peak when they're building data centers. And so that's incredible. But like their balance sheet is full. There's only so many people that you can marshal and resources to build out another gigawatt when they're going to 1.5 and beyond." "So Nebius has really capitalized on this middle tier of the market, which is not small." "Like I said, this is customers who are coming to us and saying, hey, I saw you write such good things about CoreWeave. I just have $100 million. I'm trying to give it to them. They're telling me they can't accept it until May of next year because they're backed up. And therefore, Nebius has been capitalizing." ______ For more commentary on CoreWeave and Nebius: firesidealpha.substack.com/p…

Spoke 1:1 with the most excellent @FundaAI today, and I will be doing expert interviews on their platform. Although most people prefer to keep transcripts to institutional clients, they are letting me post to my Substack, and it will become part of their expert network database!

A small personal note. I started posting much more frequently on X earlier this month — mostly takeaways from conferences, company visits, investor meetings, and conversations with IR teams. I’m still a tiny voice on X, but over the past few weeks, people I’ve followed and learned from for a long time started following me back, reaching out, sharing my posts, and telling me that some of what I shared was useful. I’m genuinely very grateful. I’ve always believed that helping others is also one of the best ways to become a better version of yourself. I think I’m slowly finding something I really enjoy doing: connecting dots between technology, capital and people. Being based in Shenzhen/Hong Kong gives me a closer view into what’s happening on the ground in China across technology, business and investing. As AI infrastructure buildout accelerates, I believe the global ecosystem will only become more interconnected. What happens in one part of the world can increasingly matter everywhere else, and there’s a lot we can learn from each other. If I can help make some of what’s happening here more visible to investors, companies and researchers around the world, even in a small way, that already means a lot to me. I’m also still very much learning. I’m not a veteran industry expert, and my coverage can’t match that of large research institutions or many great researchers on X. Most company information I share comes from normal IR, conferences and company meetings. I don’t seek or share confidential or non-public information. Many of my views are simply my own research and thinking, shaped by conversations with people I learn from. So please take what I post as part of an ongoing learning process. I’ll get things wrong, and I’m always happy to be corrected. Field research isn’t always smooth. Sometimes I’m just a random kid in the room and people don’t really want to talk to me. But that’s okay — I’ll keep showing up. Along the way, I’ve learned a lot, met great people, built meaningful connections, and realized that some of what I share can actually help others. That’s more than enough to keep me going. A bit about me: I’m from Shenzhen, studied CS in the US for both undergrad and grad school, then worked as a software engineer in the Bay Area for around three years. I’m now based in Hong Kong. If you’re in Hong Kong or Shenzhen and are interested in AI, tech, investing, or just exchanging ideas, always happy to connect :) This month has been unusually conference-heavy, so I’ll slow down a bit and take a short break. Longer term, I’ll keep sharing my own research, industry work, conversations with companies and IR teams, perspectives from investors in Mainland China and Hong Kong, and whatever else I find interesting. For longer research and thoughts, I’ll also post on Substack — everything will stay free: substack.com/@awodias Thank you to everyone who reads, interacts, shares ideas, or teaches me something.