Accelerate the development of AI industry with Chiplet technology-Blue Ocean Intelligence launches high-performance low-power AI chip

On July 9, 2021, the World Artificial Intelligence Conference, in order to accelerate the industrialization of artificial intelligence technology, as a leading large computing power chip innovation enterprise mastering the third-generation semiconductor core technology-Chiplet – Nanjing Blue Ocean Intelligent Technology Co., Ltd. in Shanghai Co-sponsored the establishment of the Artificial Intelligence Computing Industry Ecological Alliance (ICPA).

Nanjing Lanyang Intelligent Technology Co., Ltd. was founded in 2019. The company has deep technical accumulation in GPU, NPU, multimedia and high-efficiency computing. The company’s core team has more than 20 years of work experience in the above fields. The team has shipped more than 10 billion chips and has the processing capacity to ship 2 billion chips annually. Nanjing Blue Ocean Intelligent Technology Co., Ltd. is committed to developing original technologies and products. Through in-depth research and innovation on Chiplet architecture-related technologies, it is currently the first semiconductor company in the world that uses Chiplet architecture technology to develop AI chips with high computing power. The company will realize the tape-out of two chips using the 12 nm process and Chiplet architecture technology in Q4 2021. With the launch of the above two chips, the company will simultaneously release to the industry different computing power requirements such as end, edge and cloud. Chip products and solution products.

Through the leading Chiplet architecture technology and implementation, the company’s GPGPU product and solution BlueFin can support end-to-end high-precision computing (FP32) applications, and can achieve flexible and scalable solutions with different computing power from 2 TFLOPs to 128 TFLOPs ; The company’s NPU chip and solution BlueDanio supports various high computing power and energy efficient AI training and reasoning applications, and can achieve flexible solutions with different computing power from 50 TOPs to 1200 TOPs. At the same time, with the company’s integrated software and hardware development environment, it will greatly enhance the competitiveness of customers’ products, reduce product development costs and shorten product development cycles; Go beyond the existing solution products, achieve breakthroughs in high-end general-purpose intelligent computing chips, and bring higher value to customers!

At present, the company has launched cooperation projects with core partners in many different application fields. Based on the SDK provided by the company, the first batch of core partners have completed the transplantation of the original solution, and the performance of the solution has been preliminarily verified; with the help of the company’s customer support team, the core customers have also completed the unique system architecture and flexibility based on the company’s products. Prototype development of new applications with scalable functions to greatly improve system capabilities and efficiency and reduce system costs.

The Links:   FP75R12KT4_B16 PM15CSJ060 SEMIKRONIGBT

From laboratory to production line: Xilinx helps Baidu Apollo autonomous driving computing platform ACU roll… Thinklogical debuts 4K resolution KVM and video scaling solutions