https://minutesof.com/on/gpu/ Everyone on the record about gpu: 21 verbatim quotes from 4 people, February 2023 to August 2026, in one chronology, each with a timestamp and a link to… en Sun, 30 Aug 2026 22:32:59 +0000 https://minutesof.com/q/6a3bd7b5-c74b-46a6-9dd5-8db72aa9576f/ https://minutesof.com/q/6a3bd7b5-c74b-46a6-9dd5-8db72aa9576f/ “Supply of silicon can go many ways. You can either leverage it to high throughput things or high interactivity things.” — Dylan Patel, SemiAnalysis Mon, 17 Aug 2026 15:00:06 +0000 gpu https://minutesof.com/q/3047ee19-387d-4066-8826-f26d2fc2c5b3/ https://minutesof.com/q/3047ee19-387d-4066-8826-f26d2fc2c5b3/ “Every day it runs on an automated CI, and we run it on all the latest Chinese models from GLM, Zebu, Moonshot, Kimi, Alibaba, all these models we run.” — Dylan Patel, RAISE Summit Thu, 16 Jul 2026 16:44:57 +0000 gpu https://minutesof.com/q/250999e7-9869-45e6-95fa-d85ab1f347ac/ https://minutesof.com/q/250999e7-9869-45e6-95fa-d85ab1f347ac/ “It's got more more, memory, flops. It's got it's it's better than Vera Rubin in most every way. It's a little bit later. I mean, three to six months after,” — Dylan Patel, Supermicro Wed, 15 Jul 2026 00:08:28 +0000 gpu https://minutesof.com/q/ea805895-0928-4287-937a-0a6ccfad436b/ https://minutesof.com/q/ea805895-0928-4287-937a-0a6ccfad436b/ “It's 72 GPUs. It's got more memory than Vera Rubin. It's got more more, memory, flops. It's got it's it's better than Vera Rubin in most every way. It's a little bit later.” — Dylan Patel, Supermicro Wed, 15 Jul 2026 00:08:28 +0000 gpu https://minutesof.com/q/32312d2c-7043-4e82-9029-8dcbaeafc456/ https://minutesof.com/q/32312d2c-7043-4e82-9029-8dcbaeafc456/ “Tranium sells at sub $10,000,000,000 per gigawatt rental rate to Anthropic and to OpenAI. GPUs, at least before the craziness of the last six months, usually went around 12 to $13,000,000,000 per gigawatt.” — Dylan Patel, Sequoia Capital Tue, 30 Jun 2026 12:00:25 +0000 gpu https://minutesof.com/q/e48abb03-3ad3-444d-b363-f1842caaa0da/ https://minutesof.com/q/e48abb03-3ad3-444d-b363-f1842caaa0da/ “This is going to be really good for the whole private credit industry. It's going to help finance the AI build out.” — Gavin Baker, Invest Like the Best Wed, 20 May 2026 08:00:00 +0000 gpu https://minutesof.com/q/8069a2f4-c1da-4a82-b880-1c456a111132/ https://minutesof.com/q/8069a2f4-c1da-4a82-b880-1c456a111132/ “We try and track the entire supply chain from tools that manufacture chips, fabs, data centers, energy, industrials, and then AI models and who's using them, how much, and where.” — Dylan Patel, Swole as a Service Wed, 13 May 2026 20:00:36 +0000 gpu https://minutesof.com/q/267f5cee-22ee-4be6-bf05-893375b653c5/ https://minutesof.com/q/267f5cee-22ee-4be6-bf05-893375b653c5/ “When you first turn on the cluster, about ten to fifteen percent of them fail RMA in the first two weeks. Wow. And then that's fine.” — Dylan Patel, TBPN Tue, 03 Feb 2026 23:08:03 +0000 gpu https://minutesof.com/q/0dc25ccb-27dc-490b-bdd1-24ad4612e6ab/ https://minutesof.com/q/0dc25ccb-27dc-490b-bdd1-24ad4612e6ab/ “When you first turn on the cluster, about ten to fifteen percent of them fail RMA in the first two weeks. Wow. And then that's fine. Like you have to receipt them, whatever.” — Dylan Patel, TBPN Tue, 03 Feb 2026 23:08:03 +0000 gpu https://minutesof.com/q/d22ded74-0a82-4ccb-b2a1-0abee1378bda/ https://minutesof.com/q/d22ded74-0a82-4ccb-b2a1-0abee1378bda/ “Even though it's terrible on a TCO basis, they were able to get, you know, 10% more GPUs in, and it's great. And Meta has done similar.” — Dylan Patel, Open Compute Project Thu, 23 Oct 2025 06:01:45 +0000 gpu https://minutesof.com/q/b4be77a2-c887-404a-a346-9a375dad5b9f/ https://minutesof.com/q/b4be77a2-c887-404a-a346-9a375dad5b9f/ “One of the more peculiar of all the deals is, and this was disclosed in a CoreWeave filing, was NVIDIA has promised to buy any of CoreWeave's service availability that they can't sell to anyone else. That is very unusual. That's not the same as making an investment.” — Bill Gurley, Bg2 Pod Tue, 14 Oct 2025 23:56:17 +0000 gpu https://minutesof.com/q/13cb1933-7756-463f-ac46-7f8cd066bee0/ https://minutesof.com/q/13cb1933-7756-463f-ac46-7f8cd066bee0/ “That's over $10,000,000 of hardware, and then that's just one replica, then you'll have a lot of replicas and you share the caching servers between them.” — Dylan Patel, No Priors: AI, Machine Learning, Tech, & Startups Thu, 14 Aug 2025 10:01:35 +0000 gpu https://minutesof.com/q/2e9b3ee4-e698-4516-a8b0-920ae1a288c3/ https://minutesof.com/q/2e9b3ee4-e698-4516-a8b0-920ae1a288c3/ “Just in Q3, Q4 and the early part of Q1 this year, they imported a million H20s, Right?” — Dylan Patel, Special Competitive Studies Project Thu, 13 Mar 2025 16:42:15 +0000 gpu https://minutesof.com/q/08661a67-8385-4a47-bc96-ca62dcc63083/ https://minutesof.com/q/08661a67-8385-4a47-bc96-ca62dcc63083/ “Nvidia made over 4,000,000 GPUs last year, they're making over 7,000,000 this year, right? High end data center GPUs.” — Dylan Patel, Special Competitive Studies Project Thu, 13 Mar 2025 16:42:15 +0000 gpu https://minutesof.com/q/38d113f9-ba13-446c-ad37-db93f239378f/ https://minutesof.com/q/38d113f9-ba13-446c-ad37-db93f239378f/ “And I think that is a much easier goal to achieve than trying to debate on what AGI is.” — Dylan Patel, Lex Fridman Mon, 03 Feb 2025 00:12:13 +0000 gpu https://minutesof.com/q/e90226f2-0307-4f84-8256-995419be2ebc/ https://minutesof.com/q/e90226f2-0307-4f84-8256-995419be2ebc/ “GPUs have gotten 50 times faster, and the rest of the data center has only gotten four to five times faster. And that is why MFU is so low.” — Gavin Baker, Invest Like the Best Tue, 27 Aug 2024 08:00:00 +0000 gpu https://minutesof.com/q/b33bc467-808c-4c8a-b6e6-b35fc3397fb5/ https://minutesof.com/q/b33bc467-808c-4c8a-b6e6-b35fc3397fb5/ “GPUs have gotten 50 times faster, and the rest of the data center has only gotten four to five times faster.” — Gavin Baker, Invest Like the Best Tue, 27 Aug 2024 08:00:00 +0000 gpu https://minutesof.com/q/b7585dc0-5d8f-4cd8-91aa-72ee33bf55c9/ https://minutesof.com/q/b7585dc0-5d8f-4cd8-91aa-72ee33bf55c9/ “And if people aren't willing to pay you for it but your cost went up, then your gross margins get crushed” — Jensen Huang, Stripe Tue, 21 May 2024 19:46:22 +0000 gpu https://minutesof.com/q/fb13f24f-d903-4521-9be1-a668dcec5b78/ https://minutesof.com/q/fb13f24f-d903-4521-9be1-a668dcec5b78/ “call it four or five five, six months between, you know, when an order is placed and you can actually have it installed in your data center if it got worked on immediately.” — Dylan Patel, The Inside View Wed, 09 Aug 2023 15:00:23 +0000 gpu https://minutesof.com/q/fad9d07a-a3d6-43a9-9e0b-f2591b6d3829/ https://minutesof.com/q/fad9d07a-a3d6-43a9-9e0b-f2591b6d3829/ “And now with the h 100, they have 96 or 80 gigabytes of memory, but the cost per gigabyte is the same.” — Dylan Patel, Gradient Flow Thu, 02 Feb 2023 12:00:39 +0000 gpu https://minutesof.com/q/45f2aadd-215b-4a37-97e7-67d1873d78a9/ https://minutesof.com/q/45f2aadd-215b-4a37-97e7-67d1873d78a9/ “The bandwidth has not even gone up one order of magnitude. Right? Whereas flops have gone up two orders of magnitude. So so less than one order of magnitude versus two orders of magnitude increase.” — Dylan Patel, Gradient Flow Thu, 02 Feb 2023 12:00:39 +0000 gpu