ADVSUS08-BP01 Optimize privacy workload processing patterns and resource allocation for sustainability
For privacy-enhanced collaboration, advertising workloads have specific sustainability considerations for combining first and third-party customer data directly.
Implementation guidance
-
Schedule intensive privacy computations during periods of lower carbon intensity.
-
Use batch processing for data cleansing and matching operations.
-
Implement efficient data compression and formatting using formats such as Parquet.
-
Leverage AWS Graviton processors for energy-efficient computing.
-
Use serverless architectures for matching operations where possible.
-
Implement auto scaling based on actual collaboration workload patterns.
-
Configure Regional data aggregation before central processing to reduce transfer needs.
Key AWS services
-
AWS Lambda
-
AWS Graviton Processors
-
AWS Auto Scaling
Resources
