FathomX in collaboration with Hewlett Packard Enterprise has just released a white paper highlighting the collaboration on the training model for mammogram diagnosis solution which is powered by the HPE APOLLO 6500 server with NVIDIA Tesla V100 and NVIDIA Tesla A100 GPU modules.
It outlines the testing methodology and demonstrated how FathomX and its AI-assisted mammogram solution could benefit in improved performance using the NVIDIA A100 Tensor Core GPU. The objectives of the test are to ensure that AI model training could be effectively performed under the system environment and to possess better understanding on any improvement in performance introduced by the NVIDIA A100 Tensor Core GPU.
Performance in functionality is one of the tests performed to determine that the AI model training could be executed without any problem. The tests are also performed against 30,000 relevant data set of images, measuring the average processing time. In addition, the health status of the system, specifically the CPU and the GPU was also monitored.
Download a version of it at https://www.fathomx.co/publications