Case Study: PSA Group achieves 15x faster configurator performance and 95% CO2-request offload with GigaSpaces

PSA Group Automotive Modernizes their Mainframe with GigaSpaces

PSA Group (now part of Stellantis) needed to calculate and display CO2 emissions for every possible car configuration to meet EU regulations, or face heavy fines. Their web configurator relied on a mainframe limited to about 200 requests/sec with slow response times, while expected demand was an order of magnitude higher. A basic cache failed because billions of possible configurations and the lack of secondary indexing made simple key-based caching impractical.

Working with Capgemini, GigaSpaces deployed an in-memory data grid and applied option masking to dramatically reduce the configuration space, enabling dynamic multi-criteria queries without duplicating data. Deployed in 12 weeks with automatic database backup for high availability, the solution now serves over 95% of CO2 requests, handles ~2,800 requests/sec with 15–19 ms latency (vs. 200–300 ms on the mainframe), increased calculator capacity 15x, reduced mainframe load and costs, and ensured compliance with EU rules.

PSA

Frederic Warin

Managing Enterprise Architect