A company wants to reduce costs for its three-tier web architecture. Web, application, and database servers run on Amazon EC2 instances for development, test, and production environments. EC2 instances average 30% CPU utilization during peak hours and 10% during non-peak hours. Production instances run 24 hours a day. Development and test instances run at least 8 hours each day, and the company will automate stopping them when not in use. Which EC2 instance purchasing solution will meet the company's requirements MOST cost-effectively?

Reserved Instances (RIs) are ideal for the production environment because these instances run 24/7 and have a predictable, sustained workload. RIs offer significant cost savings (up to 72%) compared to On-Demand pricing for instances with steady-state usage. On-Demand Instances are suitable for development and test environments because their usage is intermittent and can be stopped when not in use. While On-Demand pricing is higher per hour, the ability to stop and start these instances means you only pay for the compute time used, making it cost-effective for unpredictable or short-term workloads. Spot Instances and Spot Blocks are generally not suitable for production workloads that require high availability and uninterrupted service, as they can be interrupted with two minutes' notice. Using RIs for development/test instances would be less cost-effective than On-Demand because RIs require a commitment, and the instances are frequently stopped.