GPU Resource Management
GPU resource management and cost optimization using Karpenter, KEDA, and DRA on EKS
2 docs tagged with "keda"
View all tagsGPU Resources · Observability · Hybrid Node · Lessons Learned
2-Tier GPU autoscaling, DCGM/vLLM monitoring, Bifrost→Bedrock Cascade Fallback, Hybrid Node on-premises integration, large MoE deployment lessons learned