Amazon Web Services has deployed UltraClusters of 100,000 Trainium2 chips, providing 65 Exaflops of aggregate compute for training next-generation frontier AI models.
Elena Rostova
Lead AI Research & Foundation Models Editor
📷 AI Chips & Infrastructure technical analysis briefing and verified research report.
AWS announced the general availability of EC2 Trn2 instances powered by custom Trainium2 chips, enabling AI research labs like Anthropic to train next-generation models with up to 4x better price-performance compared to standard GPU instances.
Trainium2 is Amazon custom designed silicon accelerator optimized for deep learning model training and inference in AWS datacenters.
Verified against 2 independent technical sources (Factual Integrity Audited)
AWS Trainium2 Clusters Scale to 100,000 Custom Silicon Nodes for Hyperscale Foundation Models
Factual consistency confirmed with technical preprints and primary documentation.