Compare · GPUs & AI chips

Instinct MI300A vs TPU v4

RANKED BY DENSE FP16/BF16 TENSOR TFLOPS PER CHIP #10 VS #18 CHECKED 06 SEPT 2026

Side by side

Verdict

Instinct MI300A ranks #10 of 20 on Dense FP16/BF16 tensor TFLOPS per chip (980.6 TFLOPS); TPU v4 ranks #18 (275 TFLOPS).

Rank 10 of 20 · leads

Instinct MI300A

Dense FP16/BF16 tensor TFLOPS per chip

980.6 TFLOPS

MakerAMD

Date2023

Verified04 Sept 2026

Evidence

AMD Instinct MI300A product page: "Peak Half Precision (FP16) Performance 980.6 TFLOPs". Structured-sparsity counterpart on the same page is 1.96 PFLOPs and was not used. Launch date on the page: 6 December 2023. 128 GB unified HBM3. Dense FP16 sits just below H100/H200 SXM (989.5 TFLOPS).

Rank 18 of 20

TPU v4

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.

Try another pair

More Instinct MI300A 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

Full GPUs & AI chips ranking Spot an error? Send a tip