Chip comparison
NVIDIA H200 Tensor Core GPU vs SambaNova SN40L
Side-by-side specs
Straight from each vendor's datasheet. Winner in bold: higher throughput, memory, or transistor count; lower power draw or price. Sparse (2:4) throughput roughly doubles the FP8/INT8 numbers shown.
| Field | NVIDIA H200 Tensor Core GPU | SambaNova SN40L |
|---|---|---|
| Process node | TSMC 4N | TSMC N5 |
| Transistors (B) | 80 | — |
| Die size | 814 mm² | — |
| TDP | 700 W | 700 W |
| Memory | 141 GB HBM3e | 1500 GB DDR5 (three-tier: 64 MB SRAM + 64 GB HBM3 + 1.5 TB DDR5) |
| Memory bandwidth | 4,800 GB/s | — |
| FP16 (dense) | 989 TFLOPS | 638 TFLOPS |
| BF16 (dense) | 989 TFLOPS | — |
| FP8 (dense) | 1,979 TFLOPS | 1,250 TFLOPS |
| INT8 (dense) | 1,979 TOPS | — |
| Form factor | SXM5 | PCIe (SN40L DataScale system) |
| Announced | 2023-11-13 | 2023-09-19 |
| Released | 2024-03-01 | 2024-01-01 |
Robots running NVIDIA H200 Tensor Core GPU
No robots publicly running it yet.
Robots running SambaNova SN40L
No robots publicly running it yet.
Data centers with NVIDIA H200 Tensor Core GPU
Data centers with SambaNova SN40L
No data centers publicly running it yet.
Common questions
NVIDIA H200 Tensor Core GPU vs SambaNova SN40L: which is faster for training?
NVIDIA H200 Tensor Core GPU has 1.58x the dense FP8 throughput of the other (NVIDIA H200 Tensor Core GPU: 1,979 TFLOPS; SambaNova SN40L: 1,250 TFLOPS). Real-world training speed also depends on memory bandwidth, interconnect, and software stack: see the spec diff for those inputs.
NVIDIA H200 Tensor Core GPU vs SambaNova SN40L: which has more memory?
SambaNova SN40L carries more HBM (NVIDIA H200 Tensor Core GPU: 141 GB HBM3e; SambaNova SN40L: 1500 GB DDR5 (three-tier: 64 MB SRAM + 64 GB HBM3 + 1.5 TB DDR5)). Memory capacity sets the largest model that fits on one chip without partitioning; memory bandwidth sets inference throughput.
NVIDIA H200 Tensor Core GPU vs SambaNova SN40L: which is more power-efficient?
Dense FP8 TFLOPS per watt: NVIDIA H200 Tensor Core GPU 2.83 (1,979 ÷ 700 W); SambaNova SN40L 1.79 (1,250 ÷ 700 W). NVIDIA H200 Tensor Core GPU wins at the die level; system-level efficiency also depends on cooling and interconnect.
NVIDIA H200 Tensor Core GPU vs SambaNova SN40L: which is more widely used?
NVIDIA H200 Tensor Core GPU: 0 robots and 4 data centers publicly running it. SambaNova SN40L: 0 robots and 0 data centers. Empty here means no operator has said publicly, not zero adoption in the market.
See also: every chip comparison · NVIDIA H200 Tensor Core GPU full page · SambaNova SN40L full page.