Huawei today announced that its latest Ascend AI supernode fully supports DeepSeek V4 models. The company made this move as the Chinese AI startup company unveiled its most anticipated artificial intelligence tech creations – V4-Flash and V4-Pro.
In a Weibo post, Huawei revealed that its Ascend AI Supernode series has consistently supported the DeepSeek models, including the latest V4-Flash and V4-Pro versions.
The company said that the collaboration between Huawei and DeepSeek‘s hardware technologies has led the Ascend SuperPod product to now support the V4 series.
According to the Chinese phone maker, the Ascend 950 SuperPod not only redefines the performance ceiling for long context interference but also helps to achieve low-latency inference of 20ms for V4-Pro and 10ms for V4-Flash models.
“Ascend has always supported DeepSeek series models. This time, through close collaboration between the two parties in core model technology, Ascend’s entire series of supernode products now support the DeepSeek V4 series.”
Following the announcement, DeepSeek said that its Ascend chip-powered V4-Pro can outperform other open-source models in the world benchmarks alongside Google Gemini Pro 3.1, which is a closed-source model.
It further mentioned that after the Ascend 950 supernodes are mass- produced and deployed in the second half of the year, the price of the Pro version will also reduce significantly.
Huawei and DeepSeek’s collaboration for the latest AI model shows that China is achieving stable growth in the technological world without US-based products. Besides, it hints that Huawei is slowly but steadily making efforts to draw a line with US rivals like Nvidia.
(Image Credits: Huawei)
The post Ascend AI supernode fully supports DeepSeek V4 model: Huawei appeared first on Huawei Central.
