According to CAST AI’s Kubernetes Cost Benchmark 2024 Report, enterprises typically utilize only 13% of provisioned CPUs and 20% of memory in clusters with 50 or more CPUs. This finding underscores the prevalence of underutilized resources in cloud environments. Businesses can reduce cloud costs while optimizing performance by understanding and addressing these inefficiencies.
The analysis by CAST AI encompasses 4,000 clusters on Amazon Web Services (AWS), Google Cloud Platform (GCP), and Microsoft Azure (Azure) from Jan. 1 to Dec. 31, 2023, before optimization by their automation platform. The report highlights several key findings, including:
- Larger clusters with 1,000 or more CPUs typically utilize only 17% of provisioned CPUs on average.
- CPU utilization rates are nearly identical between AWS and Azure, at 11%, while Google Cloud Platform (GCP) reports a slightly higher utilization of 17%.
- Memory utilization rates vary across providers, with GCP at 18%, AWS at 20%, and Azure at 22%.
- Spot instance pricing for AWS’s six most popular instances in the US-East and US-West regions (excluding Gov regions) rose by 23% between 2022 and 2023.
Leave a Reply