The JBGeorge Tech Blog

This is a duplicate of a high-performance computing blog I authored for HPE, originally published at the HPE Newsroom on June 14, 2021.

The new era of computing arrives with an increased emphasis on the importance of sustainability. Discover how HPE is delivering on its sustainability commitments with innovative, environmentally friendly HPC and AI solutions plus flexible as-a-service and financing options.

The data explosion is not slowing down. In fact, by 2025, IDC predicts that worldwide data will grow 61%, to 175 zettabytes (175 x 1021 bytes).¹ This explosion of built-in intelligence, hyper-connectivity, and data from the edge is reshaping markets, disrupting every industry, and transforming how we live and work.

Answers to some of society’s most pressing challenges across medicine, climate change, space, and more are buried in massive pools of data. As a result, everyone—from small private companies to the largest government labs—will need technology that can perform quadrillions of operations per second to turn massive amounts of data into actionable insights. These new converged workloads are creating new workflows that drive the need for real-time, extreme scale systems.

Welcome to the exascale era

This new era of computing is quickly becoming the new norm for modern business operations. As such, it it requires an extension of the power of supercomputing with the flexibility and extensibility of cloud technologies. The capabilities of these exascale systems can enable new areas of research and discovery that will support business imperatives for security, economic competitiveness, and societal good for years to come. In fact, HPC exascale technology powers research in the areas of climate study, renewable energy, new materials research that can be used in applications to more sustainable living.

Achieving these new levels of computing power proves energy intensive and calls for an approach addressing both the costs and environmental implications of this level of power consumption.

Processors are now exceeding 200W and GPUs are operating at over 300W with components continuing to get even more power hungry going forward. High-density rack configurations in the high performance computing space are moving from 20kW to 40kW—with estimates reaching up beyond 70kW per rack in 2022. With server lifecycles lasting three-to-five years, you need to choose systems and adopt cooling strategies that can stand the test of time while still performing at the highest levels.

In the quest for higher levels of performance, it can be easy to lose sight of the fact that efficient energy usage is a key enabler of the mission to accomplish more work in the same footprint for less cost. Energy efficiency of both the data center facility and the IT equipment itself is of equal concern to ensure that the majority of the energy consumed is put to good use for actual computing. In other words, the power that is delivered to the productive portion of the system is the power that matters most.

A holistic approach coupled with a strong commitment to sustainability

As leaders in the HPC industry, HPE is producing a new breed of computing systems that are more intelligently designed and efficient than ever before. Our HPC and AI solutions encompass a supercomputing architecture across compute, interconnect, software, storage, and services—delivered on premises, hybrid, or as-a-service

We are at the forefront of transforming industries with technology and enabling our customers to harness data at exascale to advance scientific discovery and solve the world’s toughest challenges. HPE is an industry leader in sustainability, delivering customer outcomes with smart, future-proofed technology. By offering as-a-service, we are expanding options to access to these powerful solutions.

With each generation of HPC technology, we strive to dramatically improve the processing power of next-generation systems while reducing energy consumption. Our fresh approach to energy efficiency is based on an ongoing promise to fundamentally rethink the way large-scale systems are designed and built. For instance, to counteract the high amount of heat generated from the computationally-demanding activities, our solutions have liquid cooling techniques to efficiently cool large-scale systems, lowering energy needs and cost of operations.

This video confirms how and why sustainability is at the heart of our commitment to developing a high-performance computing ecosystem.https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FYmoBQAnOozY&display_name=YouTube&url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DYmoBQAnOozY&image=http%3A%2F%2Fi.ytimg.com%2Fvi%2FYmoBQAnOozY%2Fhqdefault.jpg&key=b0d40caa4f094c68be7c29880b16f56e&type=text%2Fhtml&schema=youtube

Sustainable supercomputing in action

HPE and Cray have over 70 supercomputers on the Green 500 list of the world’s most energy-efficient supercomputers, with features that optimize energy effectiveness in order to do exponentially more with less.

Physical space is another major concern in the HPC community. All of our solutions are density-optimized to pack incredible performance into a small data center space. Our storage options also work to minimize space and power taken up by storage to focus the largest amount of performance output on compute. Our industry-leading designs can cut the amount of space needed to run operations by half, resulting in energy efficiencies, maximizing stability and sustainability and future-proofing customer operations.

From a software perspective, we offer customers the ability to automatically collect and analyze power metrics for all hardware—CPU, GPU, rack, chassis, nodes, rack AC, bulk DC, and CDUs. The software also supports HPE ARCS, so users can analyze the power and cooling metrics and react to preconfigured alerts such as water leakage, power supply failure, or overheating.

Additionally, we are starting to add ML AIOps capabilities to our HPC system management software which offers administrator monitoring and management of all aspects of a cluster, including power and cooling.

Supporting HPE’s transition to an everything as-a-service company, HPC offers on HPE GreenLake. Our as-a-service model is enabling customers to reduce operational inefficiencies due to low utilization rates, overprovisioning and obsolete equipment by up to 30-to-40%. The good news is, the HPC community can now also leverage the benefits of this model.

Finally, we work closely with HPE Financial Services through the Asset Upcycling Services to help customers transition from their current infrastructure with responsible retiring, removal and disposal of IT equipment, and refurbishing IT to prevent the need for disposal.

Energy efficiency and environmental footprint play a crucial role in the long-term sustainability of HPC

Improving the energy efficiency of systems solving some of the world’s most complex problems will guarantee the capacity to achieve even greater computing possibilities to drive positive social change.

At HPE, we have transformed all aspects of HPC to enhance computing power without prohibitive increases in energy that can lead to higher costs of operation and cause irrevocable environmental damage. We lead the market in energy efficiency research and power-efficient design.

Make no mistake. We are here to define the new era of computing.