At the Huawei 2018 conference (HUAWEI CONNECT 2018) held today, Deputy Chairman of Huawei, Zhijun Xu first elaborated on the AI strategy, and officially released two AI chips – Ascend 910 and Ascend 310. At the meeting, Global Times asked Xu Zhijun whether Huawei has cooperation with Microsoft. Xu Zhijun denied cooperation with Microsoft and said it is not true. Xu Zhijun also said that Huawei and Microsoft are in contact, but Microsoft’s China data center does not use Huawei’s new artificial intelligence chip solution because Huawei does not sell chips separately.
Xu Zhijun said that it has been rumored Huawei is building a cloud AI chip. So he confirmed those rumors by launching two new AI chips.
Among the two AI chips launched by Huawei, the Ascend 910 is a chip with the highest computing density on a single chip. The computing power is far superior to that of Google and Nvidia. The maximum power consumption of the Ascend 310 is only 8W, which is the ultimate high-efficiency computing low-power consumption AI chip. It comes with a semi-precision (FP16) computing power of 8TFLOPS, an integer precision (INT8) of 16TOPS, and supports a 16-channel Full HD video decoding (H.264/265).
The Ascend 910 comes with a half-precision (FP16) computing power of 256TFLOPS, integer precision (INT8) of 512TOPS, supports 128-channel Full HD video decoding (H.264/265), and the maximum power consumption is 350W.
These two chips are expected to be officially launched in the second quarter of next year. In addition, in 2019, Huawei will release three AI chips, all of which are upgraded series.
At the same time, Huawei released a new AI strategy, fully transforming and embracing AI:
- Strong investment in basic research
In the fields of computing vision, natural language processing, decision-making reasoning, etc., building data efficient (fewer data requirements), energy efficient (lower computing power and energy consumption), safe and reliable, automatic autonomous machine learning basic ability;
- Create a full-stack AI solution
Create a full-stack, independent and collaborative, full-stack solution for cloud, edge and end, providing ample, economical computing resources, easy-to-use, efficient, and full-process AI platform;
- Investment in open ecology and talent training
Globally and continuously cooperate with academia, industry, and industry partners;
- Enhanced solution
Introduce AI thinking and technology into existing products and services to achieve greater value and competitiveness;
- Internal efficiency improvement
The application AI optimizes internal management and aligns with large-scale operation scenarios, greatly improving internal operational efficiency and quality.