A Dark Matter for Astrophysics Research

By Nicole Hemsoth

May 31, 2011

Back in 2008, the Sloan Digital Sky Survey (SDSS) came to an end, leaving behind hundreds of terabytes of publicly-available data that has since been used in a range of research projects. Based on this data, researchers have been able to discover distant quasars powered by supermassive black holes in the early universe, uncover collections of sub-stellar objects, and have mapped extended mass distributions around galaxies with weak gravitational fields.

Among the diverse groups of scientists tackling problems that can now be understood using the SDSS data is a team led by Dr. Risa Wechsler from Stanford University’s Department of Physics and the SLAC National Accelerator Laboratory.

Wechsler is interested in the process of galaxy formation, the development of universal structure, and what these can tell us about the fundamental physics of the universe. Naturally, dark energy and dark matter enter the equation when one is considering galactic formation and there are few better keys to probing these concepts than data generated from the SDSS.

Just as the Sloan Digital Sky Survey presented several new data storage and computational challenges, so too do the efforts to extract meaningful discoveries. Teasing apart important information for simulations and analysis generates its own string of terabytes on top of the initial SDSS data. This creates a dark matter of its own for computer scientists as they struggle to keep pace with ever-expanding volumes that are outpacing the capability of the systems designed to handle them.

Wechsler’s team used the project’s astronomical data to make comparisons in the relative luminosity of millions of galaxies to our own Milky Way. All told, the project took images of nearly one-quarter of the sky, creating its own data challenges. The findings revealed that galaxies with two satellites that are nearby with large and small Magellanic clouds are highly unique — only about four percent of galaxies have similarities to the Milky Way.

To arrive at their conclusions, the group downloaded all of the publicly available Sloan data and began looking for satellite galaxies around the Milky Way, combing through about a million galaxies with spectroscopy to select a mere 20,000 with luminosity similar to that of our own galaxy. With these select galaxies identified, they undertook the task of mining those images for evidence of nearby fainter galaxies via a random review method. As Wechsler noted, running on the Pleiades supercomputer at NASA Ames, it took roughly 6.5 million CPU hours to run a simulation of a region of the universe done with 8 billion particles, making it one of the largest simulations that has ever been done in terms of particle numbers. She said that when you move to smaller box sizes it takes a lot more CPU time per particle because the universe is more clustered on smaller scales.

Wechsler described the two distinct pipelines required for this type of reserach. First, there’s the simulation in which researchers spend time looking for galaxies in a model universe. Wechsler told us that this simulation was done on the Pleiades machine at Ames across 10,000 CPUs. From there, the team performed an analysis of this simulation, which shows the evolution of structure formations on the piece of the universe across its entire history of almost 14 billion years — a process that involves the examination of dark matter halo histories across history. As she noted, the team was “looking for gravitationally bound clumps in that dark matter distribution; you have a distribution of matter at a given time and you want to find the peaks in that density distribution since that is where we expect galaxies to form. We were looking for those types of peas across the 200 snapshots we tool to summarize that entire 14 billion year period.”

The team needed to understand the evolutionary processes that occurred between the many billions of years captured in 200 distinct moments. This meant they had to trace the particles from one snapshot to the next in their clumps, which are called dark matter halos. Once the team found the halos, which again, are associated with galaxy formation, they did a statistical analysis that sought out anything that looked like our own Milky Way. Wechsler told is that “the volume of the simulation was comparable to the volume of the data that we were looking at. Out of the 8 million or so total clumps in our simulation we found our set of 20,000 that looked like possibilities to compare to the Milky Way. By looking for fainter things around them — and remember there are a lot more faint things than bright ones — we were looking for many, many possibilities at one time.”

The computational challenges are abundant in a project like this Wechsler said. Out of all bottlenecks, storage has been the most persistent, although she noted that as of now there are no real solutions to these problems.

Aside from bottlenecks due to the massive storage requirements, Wechsler said that the other computational challenge was that even though this project represented one of the highest resolution simulations at such a volume, they require more power. She said that although they can do larger simulation in a lower resolution, getting the full dynamic range of the calculation is critical. This simulation breaks new ground in terms of being able to simulate Magellenic cloud size objects over a large volume, but it’s still smaller than the volume that the observations are able to probe. This means that scaling this kind calculation up to the next level is a major challenge, especially as Wechsler embarks on new projects.

“Our data challenges are the same as those in many other fields that are tackling multiscale problems. We have a wide dynamic range of statistics to deal with but what did enable us to do this simulation is being able to resolve many small objects in a large volume. For this and other research projects, having a wide dynamic range of scales is crucial so some of our lessons can certainly be carried over to other fields.”

As Alex Szalay friom the Johns Hopkins University Department of Physics and Astonomy noted, this is a prime example of the kinds of big data problems that researchers in astrophysics and other fields are facing. They are, as he told us, “forced to make tradoffs when they enter the extreme scale” and need to find ways to manage both storage and CPU resources so that these tradeoffs have the least possible impact on the overall time to solutions. Dr. Szalay addressed some of the specific challenges involved in Wechsler’s project in a recent presentation called “Extreme Databases-Centric Scientific Computing.” In the presentation he addresses the new scalable architectures required for data-intensive scientific applications, looking at the databases as the root point to begin exploring new solutions.

For the dark energy survey, the team will take images of about one-eighth of the sky going back seven billion years. The large synoptic survey telescope, which is currently being built will take images of the half the sky every three days and will provide even more faintness detection, detecting the brightest stars back to a few billion years after the big bang. One goal with this is to map where everything is in order to figure out what the universe is made of. Galaxy surveys help with this research because they can map the physics to large events via simulations to understand galactic evolution.

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