Tsinghua Startup Boosts Domestic Chip Adaptation by 30x with Self-Evolving AI Infra
量子位 · wechat · 2026-08-17
Tsinghua-affiliated startup Qingming Intelligence proposes an "AI optimizing AI" infrastructure to address the fragmented ecosystem and manual tuning bottlenecks of domestic computing power.
Core Solution & Impact:
- Autonomous Optimization & Evolution: The system automatically locates bottlenecks, attempts optimizations, and accumulates experience for reuse, creating a closed loop.
- Efficiency Gains: Complex operator development reduced from two weeks to 1 hour (336x faster); full model link-up reduced from one month to 1-7 days (30x faster); some operators accelerated up to 30x.
- Trillion-Token Factory: Leverages self-evolving tech to convert domestic computing power into cost-effective Token capacity, supporting daily trillion-level demands for the Agent era.
- Commercial Adoption: Adapted dozens of domestic chips (covering over half of the mainstream market), offering Token sales and self-evolving platform services.
More from Infra
- Ajinomoto cuts supply 30%, threatening China's AI supply chain — pstAsiatech · 2026-08-17
- Cerebras accelerates RL inference: 10-hour tasks finish in 1 hour — dejavucoder · 2026-08-17
- llmfit scans your hardware to find the optimal LLMs you can actually run — mhdfaran · 2026-08-17
- Power becomes key bottleneck for AI data centers; hyperscalers may spend $700B on AI infra in 2026 — emmanuelvivier · 2026-08-17
- Hyperscalers may spend $700B on AI infrastructure in 2026 — emmanuelvivier · 2026-08-17
- $100 of used RX 580s runs Qwen 27B at 7.39 t/s on DDR3 platform — Whole_Alternative_18 · 2026-08-17