Space Computing Energy Tech Transport Science Dev Loyalty ↗ ◐ Dark mode ◎ Enable Alerts
Compare · GPUs & AI chips

Instinct MI300X vs Instinct MI250

RANKED BY DENSE FP16/BF16 TENSOR TFLOPS PER CHIP #7 VS #16 CHECKED 06 SEPT 2026
Side by side
Verdict

Instinct MI300X ranks #7 of 20 on Dense FP16/BF16 tensor TFLOPS per chip (1.3 PFLOPS); Instinct MI250 ranks #16 (362.1 TFLOPS).

Rank 07 of 20 · leads
Dense FP16/BF16 tensor TFLOPS per chip
1.3 PFLOPS
MakerAMD
Date2023
Verified04 Sept 2026
Evidence
AMD Instinct MI300X product page, per-GPU spec: "Peak Half Precision (FP16) Performance 1.3 PFLOPs" / Peak bfloat16 1.3 PFLOPs; structured-sparsity counterparts 2.61 PFLOPs. Launch date on the page: 6 December 2023. 192 GB HBM3. Same dense FP16 figure as MI325X.
Rank 16 of 20
Dense FP16/BF16 tensor TFLOPS per chip
362.1 TFLOPS
MakerAMD
Date2021
Verified04 Sept 2026
Evidence
AMD Instinct MI250 product page: "Peak Half Precision (FP16) Performance 362.1 TFLOPs" and "Peak bfloat16 362.1 TFLOPs". Launch date on the page: 8 November 2021. 128 GB HBM2e.
Try another pair
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