The FPGA/GPU cluster is a cloud-based compute infrastructure specifically designed to accelerate compute-intensive applications, such as machine learning training and inference, video and image processing, security, data analytics, bioinformatics, and financial computing. Latest state of the art acceleration technologies including Xilinx Alveo 200 FPGA and NVIDIA Tesla V100 GPU, closely coupled with server processors constitute the backbone of this cluster. The software stack consists of a complete ecosystem of machine learning frameworks, libraries as well as runtime targeting heterogeneous computing accelerators.
In this webinar, you will get a detailed walk-through of the hardware as well as the software architecture of the FPGA/GPU cluster, with instructions on how to reserve, access and develop applications targetting the different categories of nodes, namely Cerebro (2 FPGAs), Genisys (2 GPUs) and Synergy (1 FPGA and one GPU). A practical use case and a demo will be introduced, consisting of a complete machine learning training/inference flow of a convolutional neural network (CNN) architecture including:
- Training and validation on the GPU V100.
- Inference on the FPGA Alveo 200.