OpenStack + AWS, HPC(aaS) and GPUs: A Pragmatic Guide


Putting it all together makes cluster “magic” happen

By Lionel Gibbons | June 01, 2017 | GPU, OpenStack, HPC, AWS, COD




Remember the feeling you had when you first learned to drive a car? Everything was new to you. Sure, you knew what a steering wheel was, and what each pedal was for, but you had to learn how much steering input you could safely use at different speeds. Discovering just how much clutch and throttle was needed to get rolling without stalling doesn’t come easily to most people. And I’ll bet figuring out how hard (and how soon) to press the brake pedal as you came up behind another car at a stoplight created a few anxious moments for you and your instructor. But what does this have to do with managing clusters? In many ways, learning how to drive a cluster is much like learning to drive a car. At first, it can seem overwhelming, but once you master the individual skills, you can use them together to create a genuinely exciting experience.

If managing clusters is part of your job description, Bright Cluster Manager is designed to make your life easier. The more complex your cluster is, the more Bright Cluster Manager can help you. Exactly how isn’t always obvious, so our CTO, Martijn de Vries, created a presentation that he delivered at the recent GPU Technology Conference in San Jose.

He showed conference goers that with the right software, public clouds, private clouds, HPC workload managers, and GPUs, can be used together creating a flexible, powerful compute infrastructure that is easy to manage. Watch the video and see how you can become an expert cluster driver too.


Presentation Slides >