SuperMUC-NG Enables Innovative Science with ‘Best Scientific Visualization’

By Rob Farber

May 7, 2020

Ranked the 9th fastest supercomputer in the world as of the November 2019 Top500 list, SuperMUC-NG located at the Leibniz Supercomputing Centre (LRZ) is powering innovative and energy efficient science in Europe, and delivering ground-breaking visualization results. It is designed as a general-purpose system to support applications across all scientific domains – life sciences, meteorology, geophysics, and climatology, to name a few. Astrophysics has been a dominant user group on LRZ’s supercomputing systems. And most recently LRZ has made their supercomputing resources available for COVID-19 related research.

Built by Lenovo, the SuperMUC-NG system is powered by Intel Xeon Scalable processors and utilizes an Intel OmniPath fabric interconnect. Astrophysicist Hans-Thomas Janka (Scientist, Max-Plank-Institute for Astrophysics and lecturer at the Technical University of Munich) observes, “We are insatiable when it comes to data volumes and computing power.” He continues, “We get a lot of data from stars and explosions from the cosmos by measuring radiation, elementary particles or gravitational waves. But we can only really observe a few developments. Therefore, astrophysicists develop models for evaluation and calculate them with the help of mathematical and physical equations. This easily produces terabytes of data that we can only analyze or visualize with high performance computers.” [i]

Dubbed the “next geneneration” system as indicated by the “-NG” designation, the benefits of the new SuperMUC-NG system will be felt by many as LRZ supplies its high performance computing resources to German national and international research teams. LRZ is a member of the Gauss Centre for Supercomputing (GCS), which combines the three national centers, namely High Performance Computing Center Stuttgart (HLRS), Jülich Supercomputing Centre (JSC), and Leibniz Supercomputing Centre (LRZ), into Germany’s foremost supercomputing institution.[ii]

In a recent development, researchers can use the SuperMUC-NG and the infrastructure of the LRZ for COVID-19 research including the search for vaccines and therapeutics, analyzing and forecasting spread scenarios for contingency planning, as well as exploring the virus and its behavior. [iii]

Dr. Janka explains the extraordinary scientific impact of foundational turbulence calculations performed on the earlier SuperMUC clusters, “Previously, astrophysicists could only perform much smaller calculations and thus only calculate two-dimensional models. For us, SuperMUC was a gift. Even three-dimensional simulations became possible. That was a huge breakthrough for us.” [iv] His statement is based on simulated results that consumed more than 570 million core hours on the earlier clusters.

SuperMUC-NG allows even more detailed models

SuperMUC-NG significantly augments the abilities of researchers to advance the state-of-the-art in research. For example, a team of researchers led by Australian National University (ANU) Professor Christoph Federrath used the system to run the largest magneto-hydrodynamic (MHD) simulation of astrophysical turbulence ever performed. [v] In particular, the inclusion of magnetic fields made the computation twice as challenging. The details of this foundational simulation work are discussed in the 2016 article “The world’s largest turbulence simulations”.

Using Software Defined Visualization (SDVis), the performance benefits and capabilities of the SuperMUC-NG hardware made it possible for a team of experts collaborating with Luigi Iapichino and Salvatore Cielo at LRZ to visualize simulation results using static 3D grid resolutions as high as 100483. “This unprecedented resolution allows for a dynamic range of four orders of magnitude in length scale”. The resolution is extraordinary and the team was able to report that these extreme scale visualizations demonstrate an excellent quantitative characterization of the gas Mach number as a function of spatial scale (the so-called structure function) with theoretical models in both the supersonic and subsonic regimes. [vi]

This visualization work was selected as a finalist in the Supercomputing 2019 “Best Scientific Visualization” contest held in Denver, Colorado (Nov. 17-22). [vii] The LRZ YouTube video titled “Visualizing the world’s largest turbulence simulation” describes the work and illustrates the fine details and dynamics that can be seen in the LRZ simulations at this extreme resolution.

A collaborative effort

The team partnered with visualization experts at LRZ and Intel to create the scientific visualizations presented at SC19. [viii] Each snapshot required more than 23 terabytes of disk space, creating an enormous amount of data to visualize. Using the Intel OSPRay engine and VisIt, the team was able to take advantage of nearly all of SuperMUC-NG’s 6,336 nodes. [ix] The Intel OSPRay library is an open source, scalable, and portable ray tracing engine (e.g. OSPRay) that delivers interactive, high-fidelity visualizations using Intel Architecture CPUs and is part of the Intel oneAPI Rendering Toolkit.  The VisIt application is an open-source interactive parallel visualization and graphical analysis tool for viewing scientific data.

The following images illustrate some of the minute details in the LRZ turbulence data. Figure 1 shows the density of the turbulent gas, plus some velocity streamlines.  The data are explored in “slabs”, as the full box would contain too many small details.

Figure 1: Sample ray-tracing plus velocity streamlines rendering of the 10048^3 hydrodynamic simulation, using VisIt and OSPRay on SuperMUC-NG. The displayed data corresponds to a slab with a volume of 10% of the full data cube.  (Source: SC’19 proceedings)

Figure 2 reflects a first-time demonstration that the extreme resolution 11563 model can resolve the transition between the two sonic scales. The team reports, “This in turns lets scientists infer the width distribution of filamentary structures in star-forming regions and ultimately the critical density for the formation of stars.”[x]

Figure 2: Volume rendering of the gas density structures associated with the sonic scale (i.e. regions with the gas velocity transitions from supersonic to subsonic) in the MHD simulation with grid resolution of 11523

Not just limited to astrophysics, extreme-scale astrophysics turbulence simulations can also help shed light on the general nature of  turbulent flow problems, including those found in Earth-bound cases.

Capitalizing on a four-fold increase in performance

The LRZ team built and deployed a custom version of the VisIt software in order to fully utilize the high degree of parallelism (number of cores per node and large vector registers) provided by the SuperMUC-NG Intel Xeon Platinum 8174 processors. The team writes, “This version integrates the Intel OSPRay rendering engine, embodying the software defined visualization concept, optimized for CPU usage without the need of accelerators. OSPRay uses Intel Threading Building Blocks (Intel TBB) for parallel work sharing, and integrates additional features […] which are absent in the standard version of VisIt.”

LRZ reports the new Intel Xeon Scalable processors (SKX in the following graph) to deliver much improved scaling and performance compared to the SuperMUC Phase 2 Intel Xeon E5-2697 v3 processors (HSW).[xi] The improvement can be seen in the comparative scaling results reported by the LRZ visualization team when creating their visualizations included below. The introduction of OSPRay (yellow and red lines) brings a further 8x speedup with respect to older methods (blue line). (Note the log scale of the y-axis.)

Figure 3: Upper panel: VisIt-OSPRay node-level scaling behavior over MPI tasks (Intel TBB always at work). The yellow line shows the scaling for a SuperMUC-NG Intel Xeon Platinum node and, for comparison, for a SuperMUC Phase-2 Intel Xeon E5-2697 v3 node (red line). Moreover, for the best-scaling algorithm before OSPRay (kernel-based ray-casting, blue line), two scaling points are shown. Lower pane: VisIt-OSPRay strong scaling for tasks of the presented visualization, using one (yellow line) or four (red line) MPI tasks per node. The reported time to solution refers to a single snapshot of the tomography of the 10048^3 hydrodynamic simulation simulation, from 128 to 3076 nodes (i.e. one half of the whole SuperMUC-NG). Percentile confidence intervals after 20 measures are shown when significant. (Source: SC’19 proceedings)

Rob Farber is a global technology consultant and author with an extensive background in HPC, AI, and teaching. Rob can be reached at [email protected].

[i] https://www.lrz.de/presse/ereignisse/2019-12-04_Interview-Prof-Janka-_EN/

[ii] https://doku.lrz.de/display/PUBLIC/Books+with+results+on+LRZ+HPC+Systems

[iii] https://www.lrz.de/wir/newsletter/neu/

[iv] https://www.lrz.de/presse/ereignisse/2019-12-04_Interview-Prof-Janka-_EN/

[v] https://sciencenode.org/feature/How%20does%20a%20star%20form.php

[vi] https://sc19.supercomputing.org/proceedings/sci_viz/sci_viz_files/svs103s2-file1.pdf

[vii] Visualizing the world’s largest turbulence simulation

[viii] https://sciencenode.org/feature/How%20does%20a%20star%20form.php

[ix] ibid

[x] ibid

[xi] https://www.hpcwire.com/2018/09/26/germany-celebrates-launch-of-two-fastest-supercomputers/

Subscribe to HPCwire's Weekly Update!

Be the most informed person in the room! Stay ahead of the tech trends with industry updates delivered to you every week!

MLPerf Inference 4.0 Results Showcase GenAI; Nvidia Still Dominates

March 28, 2024

There were no startling surprises in the latest MLPerf Inference benchmark (4.0) results released yesterday. Two new workloads — Llama 2 and Stable Diffusion XL — were added to the benchmark suite as MLPerf continues Read more…

Q&A with Nvidia’s Chief of DGX Systems on the DGX-GB200 Rack-scale System

March 27, 2024

Pictures of Nvidia's new flagship mega-server, the DGX GB200, on the GTC show floor got favorable reactions on social media for the sheer amount of computing power it brings to artificial intelligence.  Nvidia's DGX Read more…

Call for Participation in Workshop on Potential NSF CISE Quantum Initiative

March 26, 2024

Editor’s Note: Next month there will be a workshop to discuss what a quantum initiative led by NSF’s Computer, Information Science and Engineering (CISE) directorate could entail. The details are posted below in a Ca Read more…

Waseda U. Researchers Reports New Quantum Algorithm for Speeding Optimization

March 25, 2024

Optimization problems cover a wide range of applications and are often cited as good candidates for quantum computing. However, the execution time for constrained combinatorial optimization applications on quantum device Read more…

NVLink: Faster Interconnects and Switches to Help Relieve Data Bottlenecks

March 25, 2024

Nvidia’s new Blackwell architecture may have stolen the show this week at the GPU Technology Conference in San Jose, California. But an emerging bottleneck at the network layer threatens to make bigger and brawnier pro Read more…

Who is David Blackwell?

March 22, 2024

During GTC24, co-founder and president of NVIDIA Jensen Huang unveiled the Blackwell GPU. This GPU itself is heavily optimized for AI work, boasting 192GB of HBM3E memory as well as the the ability to train 1 trillion pa Read more…

MLPerf Inference 4.0 Results Showcase GenAI; Nvidia Still Dominates

March 28, 2024

There were no startling surprises in the latest MLPerf Inference benchmark (4.0) results released yesterday. Two new workloads — Llama 2 and Stable Diffusion Read more…

Q&A with Nvidia’s Chief of DGX Systems on the DGX-GB200 Rack-scale System

March 27, 2024

Pictures of Nvidia's new flagship mega-server, the DGX GB200, on the GTC show floor got favorable reactions on social media for the sheer amount of computing po Read more…

NVLink: Faster Interconnects and Switches to Help Relieve Data Bottlenecks

March 25, 2024

Nvidia’s new Blackwell architecture may have stolen the show this week at the GPU Technology Conference in San Jose, California. But an emerging bottleneck at Read more…

Who is David Blackwell?

March 22, 2024

During GTC24, co-founder and president of NVIDIA Jensen Huang unveiled the Blackwell GPU. This GPU itself is heavily optimized for AI work, boasting 192GB of HB Read more…

Nvidia Looks to Accelerate GenAI Adoption with NIM

March 19, 2024

Today at the GPU Technology Conference, Nvidia launched a new offering aimed at helping customers quickly deploy their generative AI applications in a secure, s Read more…

The Generative AI Future Is Now, Nvidia’s Huang Says

March 19, 2024

We are in the early days of a transformative shift in how business gets done thanks to the advent of generative AI, according to Nvidia CEO and cofounder Jensen Read more…

Nvidia’s New Blackwell GPU Can Train AI Models with Trillions of Parameters

March 18, 2024

Nvidia's latest and fastest GPU, codenamed Blackwell, is here and will underpin the company's AI plans this year. The chip offers performance improvements from Read more…

Nvidia Showcases Quantum Cloud, Expanding Quantum Portfolio at GTC24

March 18, 2024

Nvidia’s barrage of quantum news at GTC24 this week includes new products, signature collaborations, and a new Nvidia Quantum Cloud for quantum developers. Wh Read more…

Alibaba Shuts Down its Quantum Computing Effort

November 30, 2023

In case you missed it, China’s e-commerce giant Alibaba has shut down its quantum computing research effort. It’s not entirely clear what drove the change. Read more…

Nvidia H100: Are 550,000 GPUs Enough for This Year?

August 17, 2023

The GPU Squeeze continues to place a premium on Nvidia H100 GPUs. In a recent Financial Times article, Nvidia reports that it expects to ship 550,000 of its lat Read more…

Shutterstock 1285747942

AMD’s Horsepower-packed MI300X GPU Beats Nvidia’s Upcoming H200

December 7, 2023

AMD and Nvidia are locked in an AI performance battle – much like the gaming GPU performance clash the companies have waged for decades. AMD has claimed it Read more…

DoD Takes a Long View of Quantum Computing

December 19, 2023

Given the large sums tied to expensive weapon systems – think $100-million-plus per F-35 fighter – it’s easy to forget the U.S. Department of Defense is a Read more…

Synopsys Eats Ansys: Does HPC Get Indigestion?

February 8, 2024

Recently, it was announced that Synopsys is buying HPC tool developer Ansys. Started in Pittsburgh, Pa., in 1970 as Swanson Analysis Systems, Inc. (SASI) by John Swanson (and eventually renamed), Ansys serves the CAE (Computer Aided Engineering)/multiphysics engineering simulation market. Read more…

Choosing the Right GPU for LLM Inference and Training

December 11, 2023

Accelerating the training and inference processes of deep learning models is crucial for unleashing their true potential and NVIDIA GPUs have emerged as a game- Read more…

Intel’s Server and PC Chip Development Will Blur After 2025

January 15, 2024

Intel's dealing with much more than chip rivals breathing down its neck; it is simultaneously integrating a bevy of new technologies such as chiplets, artificia Read more…

Baidu Exits Quantum, Closely Following Alibaba’s Earlier Move

January 5, 2024

Reuters reported this week that Baidu, China’s giant e-commerce and services provider, is exiting the quantum computing development arena. Reuters reported � Read more…

Leading Solution Providers

Contributors

Comparing NVIDIA A100 and NVIDIA L40S: Which GPU is Ideal for AI and Graphics-Intensive Workloads?

October 30, 2023

With long lead times for the NVIDIA H100 and A100 GPUs, many organizations are looking at the new NVIDIA L40S GPU, which it’s a new GPU optimized for AI and g Read more…

Shutterstock 1179408610

Google Addresses the Mysteries of Its Hypercomputer 

December 28, 2023

When Google launched its Hypercomputer earlier this month (December 2023), the first reaction was, "Say what?" It turns out that the Hypercomputer is Google's t Read more…

AMD MI3000A

How AMD May Get Across the CUDA Moat

October 5, 2023

When discussing GenAI, the term "GPU" almost always enters the conversation and the topic often moves toward performance and access. Interestingly, the word "GPU" is assumed to mean "Nvidia" products. (As an aside, the popular Nvidia hardware used in GenAI are not technically... Read more…

Shutterstock 1606064203

Meta’s Zuckerberg Puts Its AI Future in the Hands of 600,000 GPUs

January 25, 2024

In under two minutes, Meta's CEO, Mark Zuckerberg, laid out the company's AI plans, which included a plan to build an artificial intelligence system with the eq Read more…

Google Introduces ‘Hypercomputer’ to Its AI Infrastructure

December 11, 2023

Google ran out of monikers to describe its new AI system released on December 7. Supercomputer perhaps wasn't an apt description, so it settled on Hypercomputer Read more…

China Is All In on a RISC-V Future

January 8, 2024

The state of RISC-V in China was discussed in a recent report released by the Jamestown Foundation, a Washington, D.C.-based think tank. The report, entitled "E Read more…

Intel Won’t Have a Xeon Max Chip with New Emerald Rapids CPU

December 14, 2023

As expected, Intel officially announced its 5th generation Xeon server chips codenamed Emerald Rapids at an event in New York City, where the focus was really o Read more…

IBM Quantum Summit: Two New QPUs, Upgraded Qiskit, 10-year Roadmap and More

December 4, 2023

IBM kicks off its annual Quantum Summit today and will announce a broad range of advances including its much-anticipated 1121-qubit Condor QPU, a smaller 133-qu Read more…

  • arrow
  • Click Here for More Headlines
  • arrow
HPCwire