The Network as a Scientific Instrument

By Nicole Hemsoth

June 10, 2013

In June 2012, Greg Bell was named the head of the U.S. Department of Energy’s Energy Sciences Network, better known as ESnet.

Funded by the DOE Office of Science, and managed and operated by the ESnet team at Lawrence Berkeley National Laboratory, ESnet provides reliable, high-performance networking capabilities to thousands of researchers tackling many of the world’s most pressing scientific and engineering problems: finding sources of clean energy, understanding climate change, developing advanced materials, and discovering the fundamental nature of our universe. ESnet interconnects scientists at more than 40 DOE sites with experimental and computing facilities in the U.S. and abroad, and with collaborators around the world. 

Invited to give the closing keynote address at the 2012 NORDUnet conference in Oslo, Norway, Bell delivered a presentation entitled “Network as Instrument: The View from Berkeley,” in which he argued that it’s time to start thinking about research networks as instruments for discovery, not just infrastructures for service delivery. The talk struck a chord with the audience, and Bell has since been invited to give versions of the presentation at conferences in the United States and Canada. Most recently, he contributed the April 25 keynote address at the THINK Conference 2013 organized by ORION, the high-speed network linking 1.8 million researchers in Ontario, Canada.

A video of Bell giving a version of this presentation at a meeting on the genomics of energy and the environment, sponsored by the DOE Joint Genome Institute, can be found at the end of the article.

In this Q&A for HPCwire, Berkeley Lab Computing Sciences Communications Manager Jon Bashor talks with Bell about his vision, ESnet news and more.

Question: To start, can you give us a short description of ESnet?

Bell: We’re the Department of Energy’s high-performance networking facility, engineered and optimized for large-scale science. ESnet was created in 1986, making it one of the longest-operating research networks in the world.

ESnet interconnects the entire national lab system, including its supercomputer centers and dozens of large-scale user facilities. Thanks to ESnet, tens of thousands of scientists around the world can transfer data, access remote resources, and collaborate productively. 

ESnet is more than a network, though — it’s a collection of skilled and dedicated people, and a great place to work. Even though we’re located near Silicon Valley, we find it relatively easy to attract talent, because we do cutting-edge engineering in the service of scientific discovery. 

Q: In a sense, ESnet has always been at the forefront of handling Big Data, and now the rest of the community is catching up. How big is Big Data on ESnet?

Bell: Scientific data sets can be truly enormous, up to petabytes in size, and they’re growing rapidly. Sometimes we use the term “Extreme Data” to distinguish data at this scale from the Big Data you’ve read about in other contexts.

The advent of extreme-data science naturally has an impact on the amount of traffic ESnet carries. In fact, we’re growing about twice as fast as the commercial internet — our traffic doubles every 18 months. I don’t foresee this trend slowing down any time soon, because the underlying exponential drivers just keep cranking along.

Everyone reading this understands that high-performance computing changed the way large-scale science is conducted. It’s clear that data intensity will have an important impact as well. Modeling and simulation will continue to be critically important, but these tools will be supplemented by new techniques that can extract insight from complex data sets, exchanged and accessed over ultra-fast research networks.

Q: Although ESnet was created in 1986, its profile seems to have risen considerably in the past five years or so. What’s behind this?

Bell: More and more researchers are discovering that networks are critical to their science. Faster networks mean faster discovery. In addition, ESnet was lucky enough to receive significant stimulus funds a few years ago. That investment allowed us to build the world’s first 100 Gbps network at continental scale, in partnership with Internet2. We finished that project just in time: the previous-generation network was showing its age, and we were beginning to outgrow it. Our new architecture gives us lots of headroom, and the ability to develop new architectures for maximizing scientific productivity. 

We’ve also significantly ramped up our activity in the area of science engagement, partnership, and outreach. We understand that building the world’s fastest science network is not sufficient. We need to make it useful to scientists, and easy to use. That’s harder than it sounds, and we’re still developing models for helping scientists take full advantage of the “fast lanes” we’ve engineered for them. 

One final contributor is the success we’re having with applied research and innovation. This critical activity has been enhanced by our dedicated, national-scale 100 Gbps research testbed, which has supported dozens of researchers in the public and private sectors. We’re really trying to push the envelope on a range of topics — including software-defined networking, alternatives to TCP, and security models for 100 Gbps and faster networks. 

While we appreciate the recognition, it’s not really important unless it helps us advance our overall mission, which is to accelerate discovery for DOE’s Office of Science. And I do think it’s having that effect. Vendors are coming to us to ask about the unique challenges of supporting science, and our users are beginning to have much higher expectations of ESnet. These are both good developments. 

Q: You’ve also been busy. Your talk describing the network as an instrument of discovery has led to multiple invited presentations in North America and Europe — and most recently you gave a version of it as the April 25 keynote address at the THINK conference organized by ORION, the high-speed network in Ontario, Canada. What’s the gist of your presentation?

Bell: My overall goal is to inspire the audience to start thinking about networks differently. Modern research networks such as ESnet and Internet2 (and similar networks around the world) can do a lot more than most people imagine. I try to explain how certain collaborations have profited by incorporating advanced networks into their discovery processes. High-energy physics pioneered this model, and other fields are following. I make the argument that research networks such as ESnet have evolved into extensions of large-scale discovery instruments. For example, the discovery of the Higgs Boson would not have been possible without a worldwide grid computing infrastructure, interconnected by high-speed research networks. Harvey Newman at Caltech pioneered this idea years ago, and the world has finally caught up. 

In these presentations, I also give concrete advice about how people can improve networking in their own back yard. ESnet maintains a website devoted to this sort of simple, practical advice: fasterdata.es.net. If you want to start learning about how to use advanced networks more effectively, this is the place to start. It’s a very popular website, with more hits than www.es.net

Q: Why do you think the message has resonated so well in the networking community?

Bell: It’s not surprising that networkers like to hear that their work is important! But there are a couple of deeper reasons as well. In recent years, networking had become a little dull. Thanks to the challenges of extreme data (and also to the advent of software defined networking), it’s a really exciting place to be again. This new energy is very obvious at networking conferences, and in the academic research community. There are a lot of eyes on networking at the moment. 

Q: Last question: What is ESnet focusing on for the coming year? For the next five years?

Bell: Over the next five years, our challenge will be to accommodate the remarkable growth curve in DOE science traffic while simultaneously making the network useful to many more researchers. It’s hard to believe, but even with our new 100 Gbps network and access to underlying optical capacity to carry multiple terabits per second, we will begin to feel a little cramped by 2018-20. At that point, we think we’ll need to light up a new nationwide optical fiber footprint. Whatever else we do, we’re always in a mode of acceleration and growth!

In the coming year, we’ll focus on recruiting about eight new staff, most of them technical. When you consider that we now have about 40 employees, adding eight is significant. We take recruitment very seriously at ESnet. We look for people who are at the top of their game technically, but that’s not enough — they need to be flexible, great communicators, and exemplary colleagues. 

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