Open Networking Summit Europe 2018

With containers becoming the desired deployment model for a wide range of workloads such as Big Data, IoT, storage, AI, ML, SDN and NFV solutions, Container orchestration tools like Kubernetes, need to evolve to meet stringent networking and resource management requirements for optimum utilization of compute, network and storage. Such application generate large volume of data that needs to be securely transferred in and out of storage and over the network and inherently require compression and encryption which makes them computationally expensive. Offloading such operations to the hardware accelerators results in better CPU utilization and improved application performance.

This lightening talk educates the audience on how discovery, bookkeeping, scheduling and allocation of hardware accelerators is done in kubernetes using Intel QuickAssist as an example accelerator.