Moore Threads Announces MTT S5000 Support for ByteDance's Open-Source Protenix-v2 Biomolecular Structure Prediction Model
Summary
Moore Threads announced that its MTT S5000 training-and-inference accelerator card has completed full-pipeline inference support for Protenix-v2, an open-source model developed by ByteDance's Seed team. Released in April 2026, the model matches or exceeds DeepMind's AlphaFold 3 on several core benchmarks. The adaptation marks Moore Threads' completion of end-to-end engineering validation across three core AI for Science scenarios: macromolecular structure prediction, whole-genome sequence modeling, and clinical medical imaging.
Bubble analysis
This is a software-adaptation story about a domestic GPU supporting an open-source research model — a technical-ecosystem milestone with no disclosed purchase amounts, compute orders, funding, or valuations. It therefore offers almost no direct evidence on whether AI is in an investment bubble. At most it shows domestic compute continuing to land in vertical research use cases, but adoption alone is not revenue or profit.
Related signals
linked by shared tags
- Kimi K3 Joins OpenAI's Enterprise Paid Ecosystem, Driving GPU Compute Procurement Demand#gpu
- 1.821 Billion Yuan! Aoni Electronics Subsidiary Signs GPU Compute Card Procurement Contract#gpu
- Another Big Compute Order: 301189 Signs a Further RMB 1.821 Billion GPU Compute-Card Purchase Contract, Shares Have Doubled This Year#gpu
- Who Will Grab the Next Super Entry Point? The Personal AI Assistant Race: Meta, Manus and OpenAI Are Already In, ByteDance Rushes to Catch Up#bytedance
- Doubao Sets Its Sights on Travel, Hotels and Other Businesses#bytedance