eyeball FinTech’s “eyeM5 HAL Middleware” is implanted on chip, which could assist software engineers, encryption specialists, data scientists in algorithm construction, testing and optimization, and realizing high speed computing performance, and protecting data security. Because the personnel of partners can never access algorithms developed by the company to prevent leakage of know-how, business value can be raised and guaranteed. The eyeM5 HAL Middleware supports 4 major framework models, i.e., Caffe, TensorFlow, PyTorch, Xilinx Vitis, as well as the ecosystem of eyeM5 HPC 5G DataComARM-based Xilinx FPGA supercomputing server.
eyeM5 HAL Middleware Supporting Four Major AI Model Frameworks
Source: eyeball FinTech Co., Ltd.
The eyeM5 Hardware Abstract Level, HAL, middleware is built-in eyeM5 5G DataCom ARM-based Xilinx FPGA HPC server, which can support the four popular AI model frameworks, i.e., Caffe, TensorFlow, PyTorch, and Vitis.