Compare · GPUs & AI chips
TPU v4 vs TPU v5e
RANKED BY DENSE FP16/BF16 TENSOR TFLOPS PER CHIP
#18 VS #19
CHECKED 06 SEPT 2026
Side by side
Verdict
TPU v4 ranks #18 of 20 on Dense FP16/BF16 tensor TFLOPS per chip (275 TFLOPS); TPU v5e ranks #19 (197 TFLOPS).
Rank 18 of 20 · leads
Dense FP16/BF16 tensor TFLOPS per chip
275 TFLOPS
MakerGoogle
Date2022
Verified04 Sept 2026
Evidence
Google Cloud TPU v4 documentation: "Peak compute per chip 275 teraflops (bf16 or int8)". 32 GiB HBM2 per chip.Rank 19 of 20
Dense FP16/BF16 tensor TFLOPS per chip
197 TFLOPS
MakerGoogle
Date2023
Verified04 Sept 2026
Evidence
Google Cloud TPU v5e documentation, per-chip specification table: "Peak compute per chip (bf16) 197 TFLOPs". 16 GB HBM per chip.Try another pair
More TPU v4 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