I'll start from the end: I have joined @VineVenturesLP. I'm excited, a little emotional, and mostly grateful to continue living my life. This morning @BusinessInsider published an interview with me. Thank you to @meliarobin and @JamieHeller for listening and for telling my story with sensitivity and nuance. For many people, I'm Noa Argamani, who was rescued by the IDF after 246 days in Hamas’ captivity. That's part of me, and it always will be. But it's not all of me, and I'd like you to get to know me in my new role too. If you're building something, have an idea you can't stop thinking about or you’re in the Valley and looking to help the most ambitious founders in Israel succeed, I hope to be a bridge between the two ecosystems.

Relace is now also the cheapest provider of DeepSeek v4.1 Flash on OpenRouter. Agent traces are 96%+ input tokens, so GPU utilization is dominated by prefill time. However, most of those tokens are not relevant for the next response. The encoder/decoder split allows you to use a lossier, cheap submodel for prefill without affecting the performance of decode (i.e. quality of responses). It also means you need to shuttle KV between GPUs a lot, which is why they redesigned attention (again) and made the KV 4x smaller. Fully taking advantage of the theoretical gains is a deep engineering problem that DeepSeek left as an exercise for the reader.

I am selling my island in Lake Tanganyika, Tanzania. Gorgeous place, but we don’t use it enough. $7.9 million or best offer. Anyone interested should contact me directly at tim@draper.vc

Healthcare’s data problem is now fundamentally an engineering problem, and @metriport is the best technical team I’ve seen tackle it. The result is a meaningfully better customer experience everywhere they’re deployed (incl amzn + gm). Proud to have led their latest $26M round.

Neil Movva (@neilmovva) started his career at Nvidia, working on GPUs and kernels, and has an unusually deep understanding of inference, from software to chips to power. We spend a lot of time on each of those layers, how they connect, and where the important tradeoffs are. What makes this conversation special is how detailed it is (like a 401-level class), yet Neil makes it remarkably clear and easy to follow. Today he runs Sail Research, a company building infrastructure for agents to make tokens as cheap as possible. We discuss: - Latency versus throughput - Why there are no bad chips, only bad pricing - The end of kernel engineering - Buying chips and power no one else wants - New chip architectures - Nvidia lore + his contrarian view of the company - Open source and the frontier labs I learned a ton. Enjoy! TIMESTAMPS 0:00 Intro 0:38 Building a “Token Factory” 4:21 The Future of Background Agents 13:09 Nvidia and the GPU Stack 23:27 Chips, Memory, and Transformers 36:14 The Future of AI Training Data 44:32 Chip Scarcity and Compute Arbitrage 52:44 Reinventing the AI Data Center 59:01 Power and the “Scavenger Strategy” 1:10:10 Open vs. Closed AI

Today, we raised a $13M series A to solve the GPU shortage Trillions are being spent on GPU CapEx while 80% of it sits idle The capacity already exists Virtualizing GPUs unlocks it Backed by Matrix, YC, and CEAS

Allow illiquid founders to pledge shares with a loan from the government to pay tax. The loan period is long but not infinite (e.g. 10 years). The loan is non-resource: at the end of the period, the loan is either paid back in cash, or the government assumes the shares.