Organizations seek high-performance computing as a service: HPE's Trish Damkroger

For processing substantial amounts of data and performing complex calculations, analytics, simulations, and artificial intelligence (AI) at high speeds, the adoption of high-performance computing (HPC) has accelerated around the world. However, the upfront capital investment has proved prohibitive for many organizations, paving the way for HPC as a service (HPCaaS). In this model, instead of purchasing or leasing hardware and software, companies subscribe to it through a pay-as-you-go consumption model. And is being widely used across several industries for use cases as varied as computer-aided design and engineering, autonomous driving, production optimization, predictive maintenance, drug discovery, precision medicine, fraud and anomaly detection, treasury and trading analytics, IoT/smart cities, and many more.

Trish Damkroger, Chief Product Officer and SVP and GM of HPC and AI at HPE, told Business Today, “If you look at current Request for Information (RFIs) and Request for Proposals (RFPs) over the last year, increasing numbers of organizations are asking for an HPC environment to be consumed as a service. This does not necessarily mean public or private cloud but instead demonstrates a growing trend of enterprises who do not want to have or manage their own HPC data center. People want easy access to a high-performance computing capability, and there is an increasing trend to ask to be able to consume it on-demand, as a cloud-like experience, delivering a defined capability and an outcome. Essentially, HPC becomes just like all the other utility services in our life.”

HPE launched high-performance computing (HPC) as a service in December 2020 and facilitate enterprises to leverage fully managed HPC services, in their data centers or a selected colocation facility which is particularly important for data-intensive workloads to avoid issues with data gravity, latency, or data sovereignty. TVS Motor has been one of the early adopters of HPE’s HPCaaS in the country.

“We offer a cloud experience for HPC in a location determined by the customer – either in a customer’s own data center or through one of our trusted colocation partners. HPE GreenLake for HPC brings the HPC cloud experience – to apps and data everywhere: in edges, colocations, data centers, and even a traditional cloud – enabling customers to free up capital, boost operational and financial flexibility and free up talent to accelerate what is next for their organization. Customers get cloud end-to-end capabilities – for example, CAE-as-a-service, crash-simulation-as-a-service, genomics-sequence-analytics-as-a-service, structural-biology-simulation-as-a-service, or blood-flow-simulation-as-a-service - essentially production-ready and elastic for any scale,” adds Damkroger.

The ability to invest in high-performance computing was previously limited to a smaller number of companies and research institutions. However, the trend of consuming HPC as a service is growing worldwide. Hyperion Research predicts a 17.6% YoY growth from 2019 to 2024 for what they classify as HPC Cloud across all the industry verticals and government sectors.

Interestingly, HPCaaS also contributes to reducing carbon footprint because the data centers are now mostly being collocated and moved to areas where cooling requires less or no new energy. On top of that, the design of the infrastructure itself includes innovations around cooling technology and holistic power management which provide substantial savings in power consumption.