A trademark entry for an NVIDIA deep learning workstation called the DGX A100 has surfaced. The DGX A100 could feature anywhere between eight to 16 of the upcoming GA100 Ampere GPU with 8,192 CUDA cores. NVIDIA is expected to talk more about Ampere during its rescheduled GTC 2020 keynote on May 14.
https://www.notebookcheck.net/Trademark-entry-reveals-the-NVIDIA-DGX-A100-The-first-GA100-Ampere-based-deep-learning-workstation.463448.0.html