Six AI accelerators, drawn the same way
Every vendor names its parts differently — a compute unit, an Xe-core, a streaming multiprocessor, a Tensix tile — and every vendor draws its own block diagram at its own level of detail. These six are drawn to one structure, so you can click from the card down to the thing that actually does the multiply and compare like with like.
Cross-vendor matrix
The same ten questions asked of each part. Where a vendor does not publish a figure, the cell reads “—” rather than an estimate.
How to read these diagrams
Every page carries two views of the same part, because they answer different questions.
The die map comes first: every block on the die drawn at its own place on one grid. This is the view that shows what a spec sheet cannot — which edge the memory controllers sit on, how many cores one L2 has to serve, which column of tiles a memory channel feeds. Hover a block for its detail, use the filter chips to isolate one class, and click a block to open it below.
The hierarchy follows: the card opens into its die, the die into its clusters, and down to the thing that actually does the multiply. Click any block marked open ›, use the breadcrumb to come back out, press Esc to go up one level. Every view has its own URL, so a link can point straight at a single SIMD32 inside a compute unit.
One map is real and the rest are logical, and each page says which it is. Tenstorrent publishes a SoC descriptor giving every tile's NOC coordinate, so the Blackhole map is the actual grid, drawn in the vendor's own coordinates. The three GPUs publish block diagrams but no per-unit geometry, so those maps take the vendor's arrangement and lay it out for legibility — not to scale, and not to physical position. A harvested part is drawn as the full die with the count stated, because no vendor publishes which specific units a given card fused off. Figures come only from published vendor documentation and press coverage; each page lists its sources.