NVIDIA A100 tensor core GPU can achieve excellent acceleration on all scales for AI, data analysis and high-performance computing (HPC), and meet extremely severe computing challenges. As the engine of NVIDIA data center platform, NVIDIA A100 can be expanded efficiently. Thousands of A100 GPUs can be integrated into the system. NVIDIA multi instance GPU (MIG) technology can also be used to divide each NVIDIA A100 into seven independent GPU instances to accelerate the workload of various scales. The third generation tensor core technology provides accelerated support for more accuracy levels of various workloads, reducing insight and time to market.