The specific hardware specifications of the NPU are not clear, but it is known that the KL830 processor collaborates with the CPU and GPU to perform artificial intelligence (AI) operations. It can provide up to 10eTOPS@8bit processing power while consuming less than 2W of peak power and it supports a broad range of large language models with simple parameters.
Kneron CEO Liu Juncheng stated at the press conference that firms that now deploy generative AI functions and applications rely on cloud giants such as Microsoft and Amazon, as well as a huge number of chips from companies such as NVIDIA and AMD.