As Blackwell Rental Prices Jump 48% in Two Months, Quantum QPUs Emerge as a Third Compute Path
量子位 · wechat · 2026-10-19
With Nvidia Blackwell GPU cloud rental up 48% in two months and even OpenAI cutting projects over compute scarcity, a QbitAI feature examines classic compute's ceiling and quantum computing's way through it, via an interview with Xianggan Technology founder Jin Yirong.
Key points: NP-hard problems like molecular simulation and combinatorial optimization scale exponentially, beyond brute-force GPU scaling; superposition and entanglement can encode combinatorial spaces into an exponentially growing compute space. The future is "hybrid computing" — QPUs coexisting with classical supercomputing and AI compute — with near-term focus on using AI to help calibrate and optimize quantum hardware. Xianggan, spun out of BAQIS, uses a tantalum-based superconducting route: 100+ qubit chips with >100μs coherence, 99.9%+ single-qubit and 99.5%+ two-qubit fidelity; control-system revenue grew from 3M RMB (2024) to 31M RMB (2025). It aims to deliver a 100-qubit system this year, build a kilo-qubit cloud platform with API/AI-agent access, and target finance, HPC, and weather use cases.
More from Infra
- OpenJet claims Qwen3.8-27B coding runs 1.9x faster than llama.cpp on one RTX 3090 — JeremyCMorgan · 2026-10-23
- Jensen Huang lays out NVIDIA's full Vera Rubin agentic system: NVL72, MGX, CMX and RTX — rohanpaul_ai · 2026-10-23
- Jensen Huang details Vera Rubin agentic stack: NVL72, MGX, CMX and RTX — rohanpaul_ai · 2026-10-23
- ONNX to Transformers.js: running models directly in your browser — JosephJacks_ · 2026-10-23
- MiniMax M3.1 Flash free for 3 days on 0G, with 1M context and 512K output — MiniMax_AI · 2026-10-23
- "We bet on Cloudflare and it paid off": R2 and D1 at billion scale — dinasaur_404 · 2026-10-23