AI software innovator Mipsology today announced a design partnership with E-Elements, a Taiwan-based supplier of professional FPGA service training, design, and technological services. E-Elements will bundle Xilinx solutions enhanced with Mipsology’s Zebra AI inference accelerator in products and services designed for the Asian medical, robotics, and autonomous transportation industries.
“Combining the strong FPGA expertise of E-Elements with the excellence of Mipsology’s solution for Machine Learning makes a perfect match to deliver best-in-class solutions to the industry,” said Young Wang, President, of E-Elements. “Zebra will help us accelerate the time to market for our customers’ high-performance AI solutions – from data center to the Edge and embedded, with lower costs and longer lifespan than GPU-based systems.”
FPGAs are better suited than GPUs to accelerate decisioning (i.e., inferencing) for large industrial AI applications like digital imaging, industrial robotics and autonomous cars. However, FPGA programming requires the significant knowledge and expertise of specialized, hard-to-find hardware designers. Zebra eliminates the need for FPGA expertise to compute neural networks, making them as easy to use for deep learning inference acceleration as CPU/GPU. Running neural networks defined with TensorFlow or PyTorch on FPGAs would normally require considerable manual development time and effort; Zebra makes neural network deployment instant and effortless.
“E-Elements has an impressive history of delivering solutions to their customers,” said Ludovic Larzul, CEO and founder of Mipsology. “They understand requirements across a wide variety of markets in Taiwan and China, making them a prime addition to our growing ecosystem. We are already working together on systems that integrate machine learning and other functions on a single FPGA, making our position unique in allowing deployment of more robust complex solutions for high demand industries.”