Revision History
AMD Instinct MI355X · 10 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: Fixed the core precision error (dense vs sparse throughput: FP4/FP6 are 10.1 PF dense with no sparsity mode, FP8 5 PF dense, FP16 2.5 PF dense) across lead, spec table, datasheet paragraph, FP8/FP4 sections, MI350X table, training section and MI400 ratios; corrected the NVIDIA competitor columns from NVIDIA spec pages (B200 8 TB/s not 5, FP64 37 TF; B300 on 4NP with 208B transistors, 288 GB, FP64 about 1.25 TF, FP8/FP4 dense/sparse; NVLink 5 at 1.8 TB/s not NVLink 4; GB200 scale-out 400G); fixed stream processors per CU (64), I/O die topology (4 XCDs per IOD), and other errors (plus 24 minor fixes)
Version 10 · Nov 4, 2026, 02:08 PM
- Update: added an official video.
Version 9 · Nov 1, 2026, 08:07 AM
- Improve internal links (2026-10-13)
Version 7 · Oct 13, 2026, 10:20 AM
- Added 6 contextual internal links
Version 5 · Jul 23, 2026, 03:29 PM