China Closes the AI Gap, AMD's 192GB Inference Chip, and Supabase Evals
About this episode
Chinese AI models reach near-parity with US frontier labs at a fraction of the cost, AMD's Gorgon Halo chip surfaces with 192GB unified memory for local AI inference, Supabase open-sources a real-world coding agent benchmark
Chapters
Sources & Citations
Introducing Supabase Evals (Official Blog)
supabase.com
Supabase on X (Official Announcement)
x.com
Supabase Releases Evals: an Open Source Benchmark That Scores Claude Code, Codex and OpenCode on Real Supabase Tasks — MarkTechPost
www.marktechpost.com
Supabase launches open benchmark for AI coding agents building backends — RuntimeWire
runtimewire.com
AMD Ryzen AI Max+ PRO 495 "Gorgon Halo" spotted in Geekbench 7
videocardz.com
AMD Ryzen AI Max+ PRO 495 "Gorgon Halo" APU Appears with Radeon 8065S
www.techpowerup.com
AMD Ryzen AI Max+ PRO 495 "Gorgon Halo" Appears in Geekbench 7
www.eteknix.com
AMD Powers Next-Generation Agent Computers with Ryzen AI Max PRO 400 Series
www.amd.com
Chinese AI models challenge US dominance with cost-effective performance
cryptobriefing.com
DeepSeek V4 Full Release (July 2026): Pricing, Benchmarks & GPT-5.6 Comparison
macgpu.com
Stanford AI Index 2026: China narrows US lead to 2.7% while spending 23x less
thenextweb.com
Why companies are experimenting with cheaper Chinese AI models instead of OpenAI and Anthropic
fortune.com
Loading transcript…