Dell PowerEdge C4140 Deep Learning Performance Comparison - Scale-up vs. Scale - Page 42

Deep Learning Performance: Scale-up vs Scale-out 7.3 Results Showing Training Time Vs Accuracy These tests were run with 90 epochs to determine training time to achieve top-1% and top-5% accuracy. Figure 35 shows the results for the server 8X SXM2, POWEREDGE C4140-V100 in single and multinode mode, C4130-P100 in single and multi-node mode, and R740-P40. Results Highlights  The fastest training time was achieved by the system 8X SXM2 with 93% of accuracy convergence in 6.6 hours.  PowerEdge C4140 - Configuration K with SXM2 in multi-node configuration achieved 95% of accuracy convergence in 7.2 hours.  R740-P40 with 48.8hrs and 93% of accuracy convergence. Figure 35: Longest tests to extract accuracy convergence and training time Architectures & Technologies Dell EMC | Infrastructure Solutions Group 41