Huawei is expanding its artificial intelligence chip efforts with the aim of competing with NVidia. Rotating chairman David Wang announced that two chips will be released in 2027:
- 960DT: the first quarter of 2027.
- Ascend 960PR: the third quarter.
The company has developed 11 semiconductors based on UnifiedBus, which enables different chips to operate like a single large computing system. Wang said the largest “supercluster” systems support 1 million artificial intelligence processors, while more than 1,000 “supernodes” have been shipped to over 370 customers. As investments continue during the period of US export controls, the ecosystem has 5,270 active developers per month.
Why it matters
This development shows that Huawei’s move in the artificial intelligence chip field is not limited to new products, but is an effort to build a broader ecosystem that combines different processors within a shared computing structure. The systems said to support a million processors, the shipment of more than 1,000 supernodes to over 370 customers, and the figure of 5,270 active developers per month indicate that the company is seeking to grow the usage and developer sides alongside its hardware. This scale raises the question of what capacity underpins Huawei’s claim of competing with Nvidia while US export controls continue. However, the inconsistency stated in the council note regarding the chips’ release schedule requires the exact timing of the plans to be clarified separately.
Background
NVidia is not a new name in the FikirPilot archive: we published 51 articles mentioning this name in the last 90 days; the most recent was dated September 19, 2026.