“From a efficiency standpoint, Nvidia’s new-generation chips — such because the B200 and the upcoming B300 Extremely, based mostly on TSMC’s 4nm course of and outfitted with superior HBM3/3E reminiscence — have considerably widened the hole in comparison with Huawei’s 910C, which is probably going constructed on SMIC’s N+2 7nm course of (successfully 14nm) and lacks superior HBM reminiscence,” mentioned Neil Shah, companion and co-founder at Counterpoint Analysis.
Shah famous that whereas Huawei’s chip might theoretically match the older Nvidia A100 or H100 in some duties, it could require extra energy and heavy software program optimization to deal with numerous AI workloads.
“International adoption of Huawei’s 910C can be hindered by restricted developer help, ecosystem maturity, and integration challenges,” mentioned Manish Rawat, semiconductor analyst at Techinsights. “Nonetheless, it presents a viable different to Nvidia’s chips for Chinese language enterprises or these affected by geopolitical constraints, particularly as US export controls restrict entry to superior Nvidia GPUs.”
Implications on enterprise AI adoption
Regardless of not being the perfect out there, the 910C might nonetheless show viable for a lot of enterprise- and hyperscale-AI use circumstances.
It might take longer to coach fashions in comparison with US-designed chips, however for a lot of, it’s a suitable trade-off given present geopolitical and provide chain dangers.
“For enterprises – notably these working in or sourcing from China – it presents a reputable different within the face of tightening US export controls, whereas supporting home innovation ecosystems like DeepSeek and decreasing dependence on international applied sciences,” mentioned Prabhu Ram, VP of the business analysis group at Cybermedia Analysis. “Though Nvidia maintains an edge in software program maturity and vitality effectivity, Huawei’s progress displays China’s rising power in competing on the forefront of AI {hardware} within the rising AI period.”
If firms working in or sourcing from China undertake Huawei’s ecosystem, it might considerably affect their procurement methods, vendor choice, and know-how evaluations, driving larger alignment with Chinese language applied sciences.