JD Cloud to build 100k-GPU AI cluster with Moore Threads

This move signals a potential inflection point for domestic GPU adoption, moving beyond pilot projects into core infrastructure at a major cloud provider. The scale of 100k GPUs suggests confidence in Moore Threads' reliability for production workloads, not just testing. It could pressure other cloud providers to diversify their hardware sourcing, especially given supply chain considerations. The open question is whether this cluster will achieve performance parity with established solutions in real-world training and inference tasks, or if it remains a strategic, less efficient alternative. Watch for benchmark results and adoption rates from external developers using the offered computing power.