B200 vs B300 Blackwell Ultra vs MI400 vs Ascend 950DT

Pyyan / Compare / B200 vs B300 Blackwell Ultra vs MI400 vs Ascend 950DT

4 of 5

GPU & AI Chips · verified 24 Sept 2026

×B200NVIDIAcurrent

×B300 Blackwell UltraNVIDIAcurrent

×MI400AMDcurrent

×HAscend 950DTHuaweicurrent

Specification	B200	B300 Blackwell Ultra	MI400	Ascend 950DT
Summary	The current volume Blackwell part.	15 PFLOPS dense FP4.	Double NVIDIA's memory capacity.	144GB on Huawei's own HBM, and the part China is building gigawatts around.
Memory	192GB HBM3e	288GB HBM3e	432GB HBM4	144GB HiZQ 2.0
Bandwidth	—	8 TB/s	19.6 TB/s	4 TB/s
Price	$30–50K	~$40–50K (verify)	—	Above ¥250,000 (about $37K), reported
Vendor	NVIDIA	NVIDIA	AMD	Huawei
Category	GPU & AI Chips	GPU & AI Chips	GPU & AI Chips	GPU & AI Chips
Official	NVIDIA ↗	NVIDIA ↗	AMD ↗	Huawei ↗

Highlighted rows are where these differ.

Best for general training.

Best for frontier training.

Best for large-model inference.

4 TB/s memory bandwidth and a 2 TB/s interconnect, 2.5x the Ascend 910C
HiZQ 2.0 is Huawei's own memory standard rather than a licensed HBM3 or HBM3e part
Indicated price rose 20% to 50% in two months to September 2026, on a memory shortage
Successor Ascend 960DT pulled forward to Q1 2027, Huawei said on 17 September

Best for serving models where Nvidia cannot be bought.