The B1 and B2 rounds were led by Alibaba Group, PICC, Jingwei Venture Capital and Wuyuan Capital, China Life Science and Technology Innovation Fund, Mirae Asset, Cornerstone Capital and Muhua Technology and Innovation Fund. Estabished shareholders Red Dot China, Yaotu Capital and Yuanmu Capital joined the funding rounds.
The company announced its first AI-video processor, the SV102, and VA1 inference acceleration add-in card for data centers in July 2021.
Vastai said it would use the funding for the implementation of the SV100 series products in domestic and foreign markets, to increase R&D in GPU products and develop other products. However, the company has not indicated what manufacturing process technology it is targeting.
SV100 series chips offer INT8 peak performance of more than 200TOPS. They support FP16, BF16 and INT8 data types supporting multiple neural networks and diverse inference scenarios such as computer vision, video processing, natural language processing, search, and recommendation systems. It also integrates up to 64 channels of H.264/H.265/AVS2 1080p video decoding, which can be used in cloud and edge intelligence.
The VA1 inference accelerator card based on the SV100 series chip is a half-height, half-length, single-slot 75W PCIe x16 card that supports 32Gbytes of memory and PCIe 4.0 high-speed interface protocol.
The SV102 is expected to go into mass production soon, Vastai said.
“Computer vision applications, which account for more than half of China’s AI application market, require high-density video decoding with AI processing to achieve end-to-end acceleration, and we have done a lot of work in balancing the two in our design,” said John Qian, co-founder and CEO of Vastai, in a translated statement.
In the same translated statement Louis Zhang, co-founder and CTO of Vastai Technologies, said, “The SV100 series chips are based on advanced DSA architecture and achieve the highest deep learning inference performance multiple times that of data center GPUs with the same power consumption.”
Next: Supported frameworks
The VastStream development platform supports models of common deep learning framework such as TensorFlow, PyTorch and Caffe2, as well as ONNX format, with customizable AI compilers to optimize model execution on Vastai hardware.
Qian said: “In the future, our world will be surrounded by trillions of pixels. From video processing and enhancement, computer vision, AR/VR, meta-universe, to intelligent driving, everything is inseparable from the underlying chip.”
Vastai was established in Shanghai in December 2018 and has more than 300 R&D engineers located in Shanghai, Beijing, Shenzhen, Xi’an, Chengdu and Toronto, Canada.
The technical storage or access is strictly necessary for the legitimate purpose of enabling the use of a specific service explicitly requested by the subscriber or user, or for the sole purpose of carrying out the transmission of a communication over an electronic communications network.
The technical storage or access is necessary for the legitimate purpose of storing preferences that are not requested by the subscriber or user.
The technical storage or access that is used exclusively for statistical purposes.The technical storage or access that is used exclusively for anonymous statistical purposes. Without a subpoena, voluntary compliance on the part of your Internet Service Provider, or additional records from a third party, information stored or retrieved for this purpose alone cannot usually be used to identify you.
The technical storage or access is required to create user profiles to send advertising, or to track the user on a website or across several websites for similar marketing purposes.