AMDAI GPU · OAM
MI350X
AMD Instinct MI350X is a CDNA 4 data-center accelerator with 288 GB HBM3E and 8 TB/s of memory bandwidth.
Specifications
Dense throughput, no sparsity| Architecture | CDNA 4 |
|---|---|
| Form factor | OAM |
| Process | TSMC N3P |
| FP16 / BF16 tensor (dense) | 2,307.2 TFLOPS |
| FP8 tensor (dense) | 4,614.4 TFLOPS |
| FP4 tensor (dense) | 9,228.8 TFLOPS |
| Memory | 288 GB HBM3E |
| Memory bandwidth | 8 TB/s |
| Power | 1000 W |
| Launch | Jun 2025 |
AI index
Tensor × bandwidth100 × √((2,307.2 ÷ 989.4) × (8 ÷ 3.35)) = 236.0
The AI index is the geometric mean of dense 16-bit tensor throughput and memory bandwidth, each relative to the H100 SXM. It is a spec-sheet comparison, not a measured benchmark.