Amazon Redshift is AWS's cloud data warehouse for SQL analytics and AI/ML. Serverless pricing from $0.375/RPU-hour, with a $300 trial credit.
Amazon Redshift is AWS's cloud data warehouse, enabling SQL analytics and AI/ML workloads across data unified from multiple sources without customers managing the underlying infrastructure. It offers both a serverless option that scales automatically and traditional provisioned clusters for workloads with predictable resource needs.
Redshift supports a lakehouse architecture with native access to open table formats such as Apache Iceberg and Parquet, and offers zero-ETL integrations with services including Aurora, RDS, DynamoDB, Kinesis, and MSK for near-real-time data availability. It integrates with Amazon SageMaker for ML model development and Amazon Bedrock for generative AI use cases, including natural-language SQL generation. Security features include row- and column-level permissions and network isolation. AWS states its RG (Graviton-based) instances deliver roughly 2.4x faster performance than prior-generation RA3 instances at about 30% lower per-vCPU cost.
Redshift uses usage-based pricing. Serverless is billed per Redshift Processing Unit (RPU) starting at $0.375 per RPU-hour with per-second billing and no charges during idle periods; optional reservations can reduce serverless costs by up to 45% for 3-year commitments. Provisioned on-demand instances start around $0.543 per hour, with managed storage billed separately (around $0.024/GB-month in US East). New Redshift Serverless users receive a $300 credit valid for 90 days.
Redshift uses usage-based pricing. Serverless starts at $0.375 per RPU-hour with per-second billing; provisioned on-demand instances start around $0.543/hour, with storage billed separately.
New Redshift Serverless users get a $300 credit valid for 90 days; provisioned clusters may offer a two-month free trial in regions without Serverless.
Yes, it natively supports open formats like Apache Iceberg and Parquet as part of its lakehouse architecture.
Yes, it integrates with Amazon SageMaker for ML development and Amazon Bedrock for generative AI, including natural-language SQL generation.