InAccel: Benefit from Enterprise FPGA acceleration on Kubernetes using SUSE Rancher

We’ve invited InAccel, a SUSE One Partner, to author a guest blog on how you could easily deploy, manage, and benefit from FPGA acceleration in Kubernetes. ~Vince

SUSE GUEST BLOG AUTHORED BY:

Elias Koromilas, CTO and Co-founder, InAccel

Ioannis Stamelos, COO and Co-founder, InAccel

SUSE Rancher and InAccel’s FPGA Operator

SUSE Rancher is a complete software stack for teams adopting and deploying containers. It addresses the operational and security challenges of managing multiple Kubernetes clusters across any infrastructure, while providing DevOps teams with integrated tools for running containerized workloads.

SUSE Rancher not only deploys production-grade Kubernetes clusters from datacenter to cloud to the edge, but it also unites them with centralized authentication, access control and observability.

One of the main advantages is that SUSE Rancher lets you streamline cluster deployment on bare metal, private clouds, public clouds or vSphere and secure them using global security policies.

Users can use Helm or the Rancher App & Marketplace catalog to deploy and manage applications across any environment, ensuring multi-cluster consistency with a single deployment.

Until now, Rancher supported GPUs as hardware accelerator resources that can be provisioned through its platform. However, FPGAs in many cases can provide higher-throughput, lower latency or much better energy efficiency compared to GPUs. This is why FPGAs are currently available as a computing resource on AWS, Azure and other major cloud providers. It is important to note that by definition, the management of FPGA applications in Kubernetes is much more challenging compared to GPU ones, as they are based on custom hardware and not on architectures.

To simplify the management of FPGAs in Kubernetes, InAccel has developed FPGA Operator, a cloud-native method to standardize and automate the deployment of all the necessary components for provisioning FPGA-enabled Kubernetes systems. The FPGA Operator delivers a universal accelerator orchestration and monitoring layer, to automate scalability and lifecycle management of containerized FPGA applications on any Kubernetes cluster.

Figure 1. Abstracted view of the Software stack for the InAccel FPGA Operator

Figure 2. InAccel FPGA Operator for easy FPGA deployment on Kubernetes

The InAccel FPGA operator also allows cluster admins to manage their remote FPGA-powered servers the same way they manage CPU-based systems, but also regular users to target particular FPGA types and explicitly consume FPGA resources in their workloads. This makes it easy to bring up a fleet of remote systems and run accelerated applications without additional technical expertise on the ground.

InAccel has developed all the required tools that enable the seamless integration of FPGAs in Rancher-managed clusters whether they are in the cloud, across a hybrid cloud or on-prem. InAccel FPGA Operator, now also available in the SUSE Rancher Apps and Marketplace, brings FPGA management in Kubernetes to life, while end users enjoy the ease of enterprise multi-cloud deployment that Rancher provides.

How to get started with enterprise FPGA deployment in Kubernetes?

InAccel FPGA Operator is now available to be deployed through the SUSE Rancher Apps & Marketplace. For more information on how to integrate your FPGAs with Rancher check out the following tutorial: https://docs.inaccel.com/labs/rancher

Want to learn more about how InAccel and SUSE can help you solve FPGA deployment and management challenges and easily benefit from utilizing FPGAs in your Kubernetes landscape? Visit our website for more information or get in touch with the InAccel team. You could also contact your SUSE Rancher sales representative – we look forward to talking to you!

Authors:

Elias Koromilas is the CTO and co-founder of InAccel. He holds a M.Eng. in Computer Engineering from National Technical University of Athens (NTUA). He has a strong Linux background and 5+ years of experience designing, implementing, and managing cutting-edge deployment automation of FPGA resources, leveraging cloud-native technologies.

Ioannis Stamelos is the COO and co-founder of InAccel. He holds an M.Sc. in Compute engineering from National Technical University of Athens (NTUA). He has more than 5 years of experience on FPGA acceleration and deployment, DevOps optimization and Full Stack development. He is the winner of the 2017 Open Hardware contest on FPGA accelerators.