Tagged
Athena still bills ~$5/TB scanned (or Capacity Reservations by DPU). July 2026 refresh — partitions, Parquet/Iceberg, workgroups, and when provisioned capacity wins.
The standard cost optimization checklist no longer cuts it. These 8 modern strategies — from unit economics to automated Savings Plans and cost observability — reflect how engineering teams are actually managing cloud spend in 2026.
AWS Graviton processors deliver 20-40% cost savings and better performance-per-watt. Complete guide: migration path, performance benchmarks, and production deployment patterns.
Karpenter replaces Kubernetes Cluster Autoscaler with intelligent bin-packing and just-in-time node provisioning. This guide covers setup, consolidation, cost optimization, and production patterns for EKS clusters.
Amazon SageMaker automates ML training, but instance costs add up fast. This guide covers spot instances, instance selection, distributed training, and production patterns to reduce SageMaker costs by 50-70%.
Bedrock billing is not a single line item — it is a composition of model invocation costs, Knowledge Base retrieval, Agent orchestration, Guardrails evaluation, and cross-region inference profile routing. Each component has its own pricing model and its own set of cost traps.
AWS Cost Optimization Hub consolidates recommendations from Compute Optimizer, Trusted Advisor, and Cost Explorer into a single prioritized list with estimated annual savings. If you are running three separate cost review processes, this dashboard replaces all of them.
Karpenter vs Cluster Autoscaler on EKS: consolidation + flexible NodePools beat half-empty ASG nodes. July 2026: karpenter.sh/v1, WhenEmptyOrUnderutilized, Spot mix checklist.
Anonymized Series A SaaS bill: ~$40k/month for ~$8–9k of real work. July 2026 refresh — free S3/DynamoDB gateway endpoints, EBS zombies, RDS rightsizing, Trusted Advisor on Business Support+.
A healthcare estate was burning ~$200k/year on NAT. July 2026 refresh — NAT still ~$0.045/GB, free S3 gateway endpoints, Secrets Manager interface endpoints, and a VPC fix pattern that cut ~78%.
Autoscaling was supposed to make costs predictable by matching capacity to demand. Instead, it introduced feedback loops, burst amplification, and — with AI workloads — a new class of non-deterministic spend that no scaling policy anticipates.
Observability is not free, and the industry has collectively underpriced it. CloudWatch log ingestion, metrics explosion, and X-Ray trace volume can together exceed your compute bill — especially once AI workloads introduce high-cardinality telemetry at scale.