Home > AI Solutions > Gen AI > White Papers > Dell PowerScale Storage Reference Architecture for NVIDIA DGX SuperPOD > Recommender - DLRM
The recommender model has different training characteristics than ResNet-50 and BERT in that the model trains in less than a single epoch. This means that the dataset is read no more than once, and local caching of data cannot be used. In addition, the file reader uses DirectIO that stresses the file system differently than the other two files. The data are formatted into a single file.
This test is only run as a single node test; however, several tests are run where the number of simultaneous jobs vary from one to the total number of nodes available. It is expected that the shared file system only sustains performance up to the peak performance measured from the Micro Benchmark. For 20 simultaneous cases, the storage system would have to provide over 120 GiB/s of sustained read performance.