Compare · GPUs & AI chips
H200 SXM vs TPU v6e (Trillium)
RANKED BY DENSE FP16/BF16 TENSOR TFLOPS PER CHIP
#8 VS #11
CHECKED 06 SEPT 2026
Side by side
Verdict
H200 SXM ranks #8 of 20 on Dense FP16/BF16 tensor TFLOPS per chip (0.990 PFLOPS); TPU v6e (Trillium) ranks #11 (918 TFLOPS).
Rank 08 of 20 · leads
Dense FP16/BF16 tensor TFLOPS per chip
0.990 PFLOPS
MakerNVIDIA
Date2024
Verified04 Sept 2026
Evidence
NVIDIA H200 product page, H200 SXM column: FP16/BF16 Tensor Core 1,979 TFLOPS. Footnote 2: "With sparsity." Dense is 1,979 / 2 = 989.5 TFLOPS (0.990 PFLOPS). 141 GB HBM3e. Same Hopper Tensor Core throughput as H100 SXM; H200's difference is memory.Rank 11 of 20
Dense FP16/BF16 tensor TFLOPS per chip
918 TFLOPS
MakerGoogle
Date2024
Verified04 Sept 2026
Evidence
Google Cloud TPU v6e documentation, per-chip specification table: "Peak compute per chip (bf16) 918 TFLOPs". Google does not publish a separate dense FP16 cell for v6e; this row uses the official per-chip BF16 figure. 32 GB HBM per chip.Try another pair
More H200 SXM matchups
Source Vendor product pages and official spec documents (AMD Instinct MI355X / MI350X / MI325X / MI300X / MI300A / MI250X / MI250 / MI210; NVIDIA…
Last checked 06 Sept 2026