Machine learning (ML) and deep learning (DL) technologies are empowering progress and are instrumental at every stage of the growth of modern enterprises. The surging interest in machine learning can be attributed to growing volumes and varieties of available data, cheap and powerful computational processing, and affordable data storage. This means that it is possible to quickly and automatically produce models that can analyze bigger, more complex data and deliver faster, more accurate results while using increasingly cheaper hardware and software platforms.
Dell Technologies has a rich portfolio of hardware platforms that are being used to power ML and DL solutions. This paper showcases two servers from Dell Technologies that are powered by AMD™ processors and accelerated by NVIDIA GPUs; Dell EMC PowerEdge R7525 and R7515 and how they perform under various ML and DL workloads administered by the MLPerf™ Consortium. The results of these tests were submitted to the closed division/data center category of MLPerf Inference v0.7 benchmark suite and they show that these hardware platforms offer industry-leading inference performance capability, and flexibility that matches the compute requirements of AI workloads.
This paper was produced by the following:
Nicholas Wakou, Shubham Billus, Ramesh Radhakrishnan
Michael Woodside, Rakshith Vasudev, Frank Han, Dharmesh Patel