Mysterious "GPU-N" in research paper could be GH100 NVIDIA Hopper GPU with 100GB of HBM2 VRAM, 8576 CUDA Cores, and 779 TFLOPs of FP16 compute

[Redaktion]

Twitter user @Redfire75369 discovered a research paper on using low-precision GPU compute for deep learning with GPU-N. The mystery card appears to be a single-die GH100 NVIDIA Hopper GPU, offering a mammoth 779 TFLOPs of FP16 compute.

https://www.notebookcheck.net/Mysterious-GPU-N-in-research-paper-could-be-GH100-NVIDIA-Hopper-GPU-with-100GB-of-HBM2-VRAM-8576-CUDA-Cores-and-779-TFLOPs-of-FP16-compute.585187.0.html