Enabled by improvements including new instance types and fast interconnects, high-performance computing (HPC) users are increasingly shifting workloads to the cloud. With cloud usage increasing, however, managing and containing costs is a growing concern. As organizations become more reliant on cloud, they are also concerned with staying portable and flexible, and avoiding lock-in to a single cloud ecosystem or provider. 

We recently examined the case for HPC hybrid cloud and shared the results in a white paper, Managing TCO in HPC Hybrid Cloud Environments, in which we explain how operators can manage total cost of ownership (TCO) effectively. We present a simple TCO model to illustrate how Altair cloud automation and spend management solutions such as Altair Control™ and Altair Navops Launch™ can boost productivity and reduce cloud-related expenses while delivering a compelling return on investment (ROI). 

HPC in the Cloud 

Fueled by the need for fast, convenient access to large-scale computing resources, the use of HPC cloud is accelerating. After years of being stuck at under 10% growth, Hyperion Consulting estimates that HPC cloud revenue will grow at a compound annual growth rate (CAGR) of 17.6% through 2024, making cloud the fastest growing HPC segment. According to Hyperion’s 2020 market estimates, cloud revenue is now growing at 2.5x the rate of on-premises server HPC revenue. 

Across multiple industries, HPC users see advantages to using cloud computing. These advantages include fast and convenient access to specialized resources such as GPU-capable instances, the ability to quickly scale infrastructure to accommodate “surge” workloads to meet changing business requirements, and the opportunity to avoid the cost and complexity of managing on-premises infrastructure. Cloud also gives organizations the ability to move to consumption-based pay-per-use models and reduce long-term capital investments, and to utilize new artificial intelligence (AI), machine learning, and data analytics applications that require HPC infrastructure. 

The Value of HPC Cloud 

Evaluating the economics of running HPC workloads in the cloud can be challenging, but HPC in the cloud can provide strong business value. Western Digital deployed a 1-million-vCPU cluster, powered by Altair® Grid Engine® and comprised of over 40,000 AWS spot instances. As a result, they reduced the runtime for a large-scale multiphysics job comprised of 2.5 million simulation tasks from 20 days to just eight hours. Their cost was approximately half of what running the same workload in-house would have been — over $100,000 in savings. 

Altair's Managing TCO in HPC Hybrid Cloud Environments white paper takes a deep dive into getting value from cloud and hybrid cloud HPC environments, reducing TCO and increasing ROI. It explores topics including cloud business value, spend management, sophisticated cloud bursting tools, and multi-cloud environments. The simple model we present for assessing HPC cloud value estimates a three-year TCO for an on-premises HPC environment and with different hybrid cloud configurations. 

Download the white paper for detailed analysis.