Where Open Source Communities Thrive

GPUs are expensive, and most teams waste over 80% of what they pay for. Aditya Krishnakumar (Senior SRE at SentinelOne) shares a real-world playbook for combining the NVIDIA GPU Operator, Dynamic Resource Allocation (DRA), and GPU time-slicing to run bursty AI workloads on Kubernetes at a fraction of the cost — achieving 65–70% lower GPU spend, 4× utilization, and flat p95 latency.