Skip to main content

Exascale

Exascale means massive computational power. Think of a system capable of a quintillion calculations per second – that’s the power of exascale.

What is exascale?

Siemens is at the forefront of exascale, exploring the next level of high-performance computing (HPC). We’re pioneering tools that scale, run on next-generation systems and support increasingly complex HPC requirements in areas like machine learning, deep learning and multiphysics.

Two people waling through a hallway looing at electronics.

Uncover the benefits

Research and innovation – faster than ever

The power of exascale is scaling up discovery and speeding innovation. Exascale computing enables new insights in weather prediction, climate modeling, drug discovery, precision medicine, aeronautics, space science and more.

HPC – only as powerful as the software that runs it

Powerful computing hardware requires powerful software, and it’s critical to manage exascale workloads efficiently. HPCWorks PBS Professional orchestrates it all with automated job scheduling, management, monitoring and reporting.

Making history with Argonne’s Aurora supercomputer

The Aurora exascale system at Argonne National Laboratory, with workload management by HPCWorks PBS Professional, can perform two quintillion — a billion billion, visualized as 1,000,000,000,000,000,000 — calculations per second.

Discover key features of exascale

trends-exascale-next-big-thing-image-right-interior-desktop-500pxhigh-jpg

Pioneering exascale

We’re pioneering scalable tools for the exascale era that address complex HPC needs in AI, machine learning, deep learning and multiphysics. As an ECP Industry Council member, Siemens helps shape the U.S. exascale ecosystem, advancing precision medicine, additive manufacturing and biofuels.

trends-exacale-going-big-image-right-interior-desktop-400pxhigh-jpg

Large-scale workload management

HPCWorks PBS Professional runs big and runs fast, orchestrating 50,000-node clusters, 10M jobs in a queue, 1,000 concurrent active users, 10M jobs per hour end-to-end throughput and a 10-second end-to-end run. It powers efficient workload management on the world's biggest computing systems.

Optimizing exascale computing

Exascale demands speed and scale to handle billions of operations. It requires extreme resilience and continuous fault recovery. Running an exascale-class system takes significant power, so power management – and a good user experience – is paramount. We’ve considered all this and more.

  • Dual-licensed HPCWorks PBS Professional with an extensive plugin framework
  • HPCWorks Velocity Plus for 6-10x greater throughput
  • PMIx to quickly launch huge MPI jobs
  • Active-active HPCWorks PBS Professional server
  • Hybrid cloud capabilities, ARM64, GPUs and SSDs
  • Security focus

Explore exascale related products

Learn more

Workload management for exascale computing

Drive breakthroughs using industry-leading workload management with top exascale systems.

Simulating realistic supernova behavior in 3D

University researchers use Argonne HPC resources and HPCWorks to advance space science.

Improving fidelity scaling

Argonne researchers use HPCWorks to smooth out quantum molecular dynamics simulations.