Revision History
AWS Inferentia · 8 revisions
Sizes are character counts of the article source. The signed number is the change from the previous revision.
Recent edit summaries
Detailed summaries recorded by editors. Generic maintenance summaries are omitted here; every recorded revision remains below. Dates describe the edit, not necessarily the event it covers.
- Correction: Corrected the Trn2 UltraServer chip count (64, not 16), the false claim that Inferentia1 had no chip-to-chip interconnect, the scope of AWS's 2.3x/70% claim (Inf1 only) and 50% performance-per-watt claim (versus other EC2 instances, not Inf1), the Amazon Search case study, and unsupported statements about Claude training and inference hardware (plus 13 minor fixes)
Version 8 · Nov 5, 2026, 09:05 AM
- Update: added an official video.
Version 7 · Nov 1, 2026, 08:10 AM
- Date the Inf1 price sentence to the paragraph's early-2026 survey
Version 6 · Oct 15, 2026, 05:58 PM
- Improve internal links (2026-10-13)
Version 5 · Oct 13, 2026, 10:04 AM
- Added 6 contextual internal links
Version 4 · Jul 23, 2026, 03:31 PM