Reported impact
Business outcome
Alibaba Cloud gave Dadi Quantum an elastic foundation for commercial AI forecasting instead of forcing the company to overbuild local infrastructure. The platform can train large weather models with hundreds of CPUs and GPUs, keep vast historical datasets in lower-cost storage tiers, and deliver more precise 15-day weather and near-real-time power predictions that help renewable-energy customers improve planning, reduce grid-connection losses, and respond more confidently to electricity-market volatility.
- Cost Reduction
- Risk Reduction
- Scalability
Solution
Dadi Quantum built its AI remote-sensing stack on Alibaba Cloud so it could convert satellite, drone, and sensor data into weather and renewable-power forecasting services without carrying fixed infrastructure sized for peak training runs. Multi-source spatial data from NASA, the European Space Agency, and other providers is ingested into the company's cloud-native spatiotemporal PaaS, where proprietary algorithms grid, index, and organize the data pool for downstream model development and SDK access. On top of that data layer, Dadi Quantum trains end-to-end AI meteorology models for short-term weather prediction and site-level power forecasting, using Alibaba Cloud GPU instances for burst compute, CPFS for high-throughput low-latency training reads, and OSS for tiered long-term storage of massive historical datasets. Forecast outputs then feed renewable-energy planning, dispatch, and trading scenarios through the platform's SDK and application services.
Data flow
Satellite, drone, and sensor feeds are ingested into Dadi Quantum's spatiotemporal PaaS, where proprietary algorithms grid and organize the data pool. AI meteorology models train on that corpus using GPU instances, CPFS, and OSS, and the resulting weather and power-forecast outputs are delivered through SDK and platform services to renewable-energy and other industry users.
