MLPerf Meets Kubernetes

Explore the intersection of MLPerf and Kubernetes in this 32-minute conference talk by Xinyuan Huang and Elvira Dzhuraeva from Cisco. Gain insights into adapting MLPerf, a state-of-the-art benchmark suite, to evaluate machine learning performance on Kubeflow and Kubernetes. Learn how Kubeflow is evolving as a cloud-native Machine Learning platform and discover how MLPerf results can guide the future design and optimization of cloud-native ML platforms. Understand the importance of performance evaluation in real-world ML workloads, covering both training and inference phases of the ML lifecycle. Delve into the potential of MLPerf to set industry standards for end-to-end performance evaluation of ML systems on Kubernetes-based platforms.