Posted in | News | Biofuels | Climate Change

EMSL Supercomputer to Aid Research in Climate and Biology-Based Fuels

A new supercomputer expected to rank among the world’s fastest machines will be ready to run computationally intense climate and biological simulations along with other scientific programs this summer. This computational work will aid research in climate and environmental science, chemical processes, biology-based fuels that can replace fossil fuels, new materials for energy applications and more.

Researchers the world over come to PNNL to use EMSL resources, including its supercomputer. EMSL/PNNL

Chosen by a competitive process, Atipa Technologies in Lawrence, Kan., will provide the machine to EMSL, the Department of Energy's Environmental Molecular Sciences Laboratory. EMSL is a national user facility on DOE's Pacific Northwest National Laboratory's campus that provides experimental and high performance computing capabilities to enable users to address environmental and energy challenges through molecular-level theory and experiment. It is also home to the new supercomputer’s predecessor, Chinook. As a national user facility resource, the new system will be available to scientists everywhere, who will be able to apply on a competitive basis to use it. Currently, about 400 scientists use Chinook.

"We're developing a supercomputer that will aid energy, environment and basic science missions important to DOE," said PNNL computational scientist Bill Shelton, the associate director at EMSL who manages high performance computing. "Enhanced computing power will benefit our users who conduct experiments and want to verify them with modeling. Integrating computational theory with experiment is critical to accelerating scientific discovery."

Funded by DOE's Office of Science, the new $17 million machine will likely peak at 3.4 quadrillion -- 3.4 million billion -- calculations per second and be more than 20 times faster than the four-year-old Chinook. The new supercomputer's capacity and speed are expected to rank it among the world's top 20 fastest machines when it comes online. Peaking at 3.4 petaflops, the new computer will be able to do in one hour what would take a typical laptop more than 20 years to do.

Atipa Technologies has been providing high performance computers to DOE and its labs for more than a decade.

"We’re excited to have the opportunity to provide the new supercomputer with a theoretical peak performance of 3.38 petaflops and 2.7 petabytes of usable storage. It will be built and deployed by Atipa Technologies in collaboration with Supermicro,” said Mike Zheng, president of Atipa Technologies.

As EMSL's flagship high performance computer, researchers from around the world will be able to use it. The EMSL team designed it for researchers who typically need resources of this scale but don't generally have access to such a powerful computer. This wide availability makes it stand out from other supercomputers.

"Its uniqueness is that it will be optimally configured for climate and chemistry simulations and biological analyses," said Shelton.

For example, the new machine will offer added speed for improved climate models. "The new computer provides a wonderful opportunity for climate scientists to get more work done and get each simulation done more quickly," said PNNL climate scientist Phil Rasch. "It is a huge jump in the computing power available to us."

And it will produce more details about how organisms work. "I'm excited because with the amount of data researchers are generating in biology, this supercomputer will open up new avenues for our users," said EMSL biology science lead Scott Baker. "More computing power is like having more pixels in a picture. We'll be able to look at proteins and complex biological interactions more realistically. This will allow us to better understand and control organisms like microbes so that we can develop new renewable fuels."

The design's 196,000 processing units are Intel processors combined with Intel Phi many integrated core (MIC) accelerator cards. The accelerator cards will ratchet up the power. They work with the conventional processors and memory and allow up to 120 extra calculations per node to be performed simultaneously rather than one at a time. (Anyone with a graphics card in their personal computer has taken advantage of a hardware accelerator.)

The system's 23,000 Intel processors have 184,000 gigabytes (184,000 billion bytes) of memory available, about four times as much memory per processor as other supercomputers. The additional memory will allow scientists to use the processors more efficiently for biology, climate research, chemistry and materials science.

Atipa will deliver the computer's components by July 2013 and assemble it at EMSL. The EMSL team will spend a few months installing and configuring the system and getting it up to speed. They expect to have it running for national and international researchers in October 2013. In the meantime, EMSL will be sponsoring a naming contest among EMSL users and friends.

The New Supercomputer's Fast Facts:
Theoretical peak processing speed of 3.4 petaflops
42 racks
195,840 cores
1440 compute nodes with conventional processors and Intel Xeon Phi "MIC" accelerators
128 GB memory per node
FDR Infiniband network
2.7 petabyte shared parallel filesystem (60 gigabytes per second read/write)
Working reference: HPCS4A

Source: http://www.atipa.com/

Tell Us What You Think

Do you have a review, update or anything you would like to add to this news story?

Leave your feedback
Your comment type
Submit

While we only use edited and approved content for Azthena answers, it may on occasions provide incorrect responses. Please confirm any data provided with the related suppliers or authors. We do not provide medical advice, if you search for medical information you must always consult a medical professional before acting on any information provided.

Your questions, but not your email details will be shared with OpenAI and retained for 30 days in accordance with their privacy principles.

Please do not ask questions that use sensitive or confidential information.

Read the full Terms & Conditions.