Amazon Ridgeland is an operational AI data center hosting an estimated 171k H100-equivalent chips, scaling toward a projected 788k H100-equivalent chips — supported by 228 MW of IT power, rising to 840 MW — at a capital cost of $8.6 B, rising to $31.8 B. We estimate that it uses Amazon Trainium2 chips. The timeline of IT power and compute capacity is based on satellite imagery, regulatory filings, chip efficiency, and third-party reporting. This site is likely part of the Anthropic–Amazon Project Rainier collaboration.