aiqbench analysis
The article covers an AI-based ensemble weather model that has been approved for operational use and trained on a 100,000-accelerator cluster, with claimed gains over ECMWF's AI ensemble system. For AI readers, the key points are that it is a deployed system and that its 11-day global range and performance claims come from the developers. Independent verification against established benchmarks would matter more than the stated comparison.
With an 11-Day Forecast Range, Supercomputing Internet Helps "Fengqing Ensemble" AI Enter Operational Service
According to 36Kr, the "Fengqing" AI global short- to medium-range ensemble weather forecasting system V1.0 recently passed operational review and was put into service. The system was trained entirely on the 100,000-card AI4S computing cluster at a core node of the National Supercomputing Internet, making it the first meteorological AI ensemble forecasting model in China for operational use with fully self-controlled data, models, and computing power. Its effective global forecast range reaches 11 days, and its overall forecasting performance surpasses that of the artificial intelligence ensemble forecasting system of the European Centre for Medium-Range Weather Forecasts (ECMWF).
Read the full article at 36KrRelated coverage

NVIDIA and Partners Build in America, for America
NVIDIA · Aug 5, 2026

Aurora 1.5: Extending open foundation models for weather and Earth-system applications
Microsoft Research · Jul 9, 2026
News Archive 2026
Quantum Computing Report · Jul 25, 2026
.jpg&w=320)
IonQ Posts Q1 2026 Earnings with Record Revenue
IonQ · May 11, 2026