NVIDIA‘s GH100 GPU uses the Hopper architecture and is made using a 5 nm production process at TSMC. With a die size of 814 mm² and a transistor count of 80,000 million it is a very big chip. GH100 does not support DirectX. For GPU compute applications, OpenCL version 3.0 and CUDA 9.0 can be used. It features 18432 shading units, 576 texture mapping units and 24 ROPs. Also included are 576 tensor cores which help improve the speed of machine learning applications.