Dell EMC Engineers Strategy to Democratize HPC

By Tiffany Trader

September 29, 2016

The freshly minted Dell EMC division of Dell Technologies is on a mission to take HPC mainstream with a strategy that hinges on engineered solutions, beginning with a focus on three industry verticals: manufacturing, research and life sciences.

“Unlike traditional HPC where everybody bought parts, assembled parts and ran the workloads and did iterative engineering, we want folks to focus on time to innovation and let us worry about the infrastructure,” said Jim Ganthier, senior vice president, validated solutions organization at Dell EMC Converged Platforms Solution Division.

The Engineered Solutions Labs at Dell’s Parmer Campus just outside Austin, as well as a new web-based system configurator aimed at streamlining the ordering and deployment process, factor prominently in Dell EMC’s plan to make HPC more accessible. That automated configurator is called System Builder and we’ll come back to how it works in a bit.

Earlier this month – the same week that the Dell-EMC merger became official — HPCwire paid a visit to the Dell EMC HPC Innovation Lab, part of the company’s Engineered Solutions Labs. The 13,000 sq. ft. shared facility houses over 1,000 servers, a range of generations and form factors targeting not only high-performance computing, but also private cloud solutions and HANA in-memory analytics workloads.

dell-emc-hpc-innovation-lab-1200x
Dell EMC Engineered Solutions Labs (Source: Dell EMC)

The HPC Innovation Lab is where Dell EMC engineers bring together compute, storage and networking technologies to build end-to-end HPC solutions. The lab’s HPC gear is focused on Dell’s spotlight verticals as well as more traditional HPC users from the government and academic arena. While some of the lab testbeds are still under wraps at this time, Dell is working with the latest technology from Intel, Nvidia, and Mellanox to understand and characterize the performance impact they have as part of a Dell EMC solution.

“We want to give our customers, our partners, our ISVs the latest and greatest access to all the technologies whether it’s Knights Landing, Omni-Path, Mellanox, Nvidia, Bright Computing, and others,” said Ganthier. “This is a chance for them to bring their workloads, play in our sandbox and in essence tune it long before it’s readily available.”

Naturally Isilon and other EMC-acquired storage technologies will also be featured prominently. Isilon is especially important in the bioinformatics sphere. “Bringing EMC Isilon together with the Dell HPC System for Life Sciences gives us a killer solution for that area,” said Ganthier.

Domain expertise is a priority for Dell EMC. “We have an engineer who worked for years at Boeing doing engine designs; he is working on our manufacturing configurations,” said Onur Celebioglu, HPC Engineering Director at Dell EMC and head of the Innovation Lab. “We have another engineer with a Masters in bioinformatics and a PhD in computer science who is working on our life sciences system for genomics. And we have engineers with computer science backgrounds, providing expertise in file systems, interconnects, and HPC management tools.”

“HPC customers being very technically oriented, they really appreciate that one-to-one engineer-to-engineer interaction,” Celebioglu continued. “We bring customers in this lab when they come visit Austin – and we let our engineers have direct interaction with our customers so that we can learn from what their needs and problems are and so they can understand the expertise that we have put in to solve their problems.”

In keeping with its specific vertical focus, the lab is home to a life sciences testbed, the Genomics Data Analysis Platform, and a manufacturing testbed, the Dell HPC System for Manufacturing, which Dell uses to PoC and optimize customer codes and to design reference architectures.

dell-genomics-data-analysis-platform-800xDell built the bioinformatics system in collaboration with its customer, TGEN, who was running a workload to analyze genome data to come up with customized treatment options for a rare form of pediatric cancer called neuroblastoma. Moving from workstations to a Dell HPC infrastructure, TGEN reduced its runtime from weeks down to just four hours per patient, said Dell EMC.

The platform is comprised of 40 Dell PowerEdge FC430 nodes, interconnected with InfiniBand. On top of the compute nodes are a parallel file system configuration with Lustre, an HPC-optimized NFS configuration and all the infrastructure nodes.

After Dell built and designed this testbed, it ran human genome pipelines in order to characterize the performance of the configuration. Lab staff also ran plants and animal genome pipelines and wrote a white paper describing the configuration and performance and power consumption metrics in a domain specific way.

“If you look at the specs of the system, it’s not provided in gigaflops or gigabytes per second,” said Celebioglu. “We try to spec this out in genomes you can analyze per day and or genomes you can analyze per watt and find the breaking point as to when do you need a parallel file system, when do you need a fast interconnect, when do you need to add other infrastructure nodes, so that we can come up with guidelines based on their needs of what size system they need to deploy for a given performance metric. That gets translated into our System Builder tool so our sales teams can go into that tool and say I want to design a system for 20 genomes per day and can come up with the right size configuration for that metric.”

dell-hpc-system-for-manufacturing-800xDell is doing the same thing for manufacturing. “We are coming up with fine-tuned systems for manufacturing applications to be able to do CFD/CAE analysis,” said Celebioglu. “Our approach here is more modular in that we are defining node types for implicit solvers and explicit solvers – one is more for crash simulation and CFD and the other is more for vibration testing and drop testing, and then we are combining those different modules together based on how many explicit solver customer needs how many implicit solver the customer needs.

“To come up with the design for those modular building blocks, we had to run a lot of these applications in-house, like ANSYS, LSTC [Livermore Software Technology Corporation], Nastran, try different processor SKUs, try different memory configurations, try different storage configurations and come up with the right components for those applications – and those went into our reference designs and system architectures.”

Both the life sciences and the manufacturing testbeds are straight x86 systems, but there are GPU-and Phi-enabled node options in the lab as well. Dell reports being one of the first Intel partners to have Knights Landing parts on site and they have an Omni-Path installation as well.

“We have done tests both with GPUs and CPUs,” said Celebioglu. “Most of the next-gen sequencing applications use straight Xeon today but there are some very interesting application use cases in the molecular dynamics space. A lot of applications are ported to GPUs – and the Knights Landing processor looks interesting for those applications as well. We keep an open mind and we’re doing a lot of tests constantly to pick the right node types for the right applications.”

All this experience and tuning has culminated into a new way of doing requirements gathering with a web-based configurator called System Builder. The modules – currently there are two, one for life sciences and one for research, with a manufacturing version planned for Q4 of this year – require the user to answer between 10 and 17 qualifying questions to get a baseline configuration.

Dell’s intention is for System Builder to get the customer to 90-95 percent and have its solutions architects take them the rest of the way, but Dell has already had one customer (the first to use the configurator) take the order as-is.

“We had a North American university whose cluster was down and they needed to act quickly,” Ganthier explained. “An enterprising Dell salesperson who had the logins to System Builder shared it with the customer. That customer answered all the questions and not only took the configured system 100 percent as-is, but that system was delivered to them in record time.”

To be clear, the target customer for System Builder is the small to mid-range user, whose mission is not computer science focused. It’s not for large HPC sites with 1,000+ nodes.

The research module poses pretty standard questions about processor type, interconnect type, number of nodes/FLOPS, etc., but the life sciences module takes a more simplified approach, focusing more on the nature of the workload and business-outcome oriented questions and less on system specifics.

The very first question it asks is: “How many human genomes per day do you want to process?”

dell-emc-system-builder-for-life-sciences-screenshot
Dell EMC System Builder for Life Sciences screenshot

The next question asks if the system is intended for general purpose? A prompt provides guidance: “In addition to NGS data analysis pipelines, if the system will be used to run workloads requiring larger memory and higher CPU power (such as simulations or statistical analyses using SAS/R), then this should be set to ‘Yes’.”

“Since these are based on standard building blocks, we custom pick the components to best fit certain workloads but essentially you can run different workloads on the system – a lot of the genomics systems are also used to run statistical analysis packages like R and molecular dynamics packages,” Celebioglu explained.

Using System Builder feels a lot like shopping online but without the upfront pricing information. The six step process results in a unique solution ID number, which begins the normal sales cycle.

“This is very well tied in to our order input system,” said Celebioglu. “Once System Builder does all the calculation it outputs a suggested configuration which is a jumping off point for further discussion.”

“There’s a whole optimization process that takes place,” said Dell EMC HPC Strategist Ed Turkel. “It’s a little unusual to have that case of the customer that purchases exactly what comes out of System Builder. They may want to change the cables or the number of racks. Or they may want to compare the performance versus price tradeoffs for different processors.”

While much of the benefit of System Builder is aimed at streamlining the configuration and ordering process, Dell EMC is also emphasizing the last mile benefits.

“What arrives at the customer site is a configured system,” said Turkel. “It’s not 200 boxes that show up with instructions and a screwdriver and a wrench. The process is optimized at both ends.”

Dell has publicly announced three HPC System Builder modules and hinted of other versions to come.

“We want to disrupt multiple industries,” said Ganthier. “It’s not just about building the product; it’s about our ability to do advisory, consulting, access to experts and our innovation sandbox as well as providing lifecycle support.”

Dell/EMC Figure 2
This slide from the 2016 HPC User Forum in Austin, Texas (Sept. 7, 2016) depicts the other high-impact verticals that Dell is targeting; presented by Dell EMC HPC Strategist Ed Turkel

 

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