A new era needs new HPC/AI storage

October 30, 2019

New ClusterStor E1000 HPC/AI storage system from Cray, a Hewlett Packard Enterprise company

Introduction / Executive Summary

High-performance computing (HPC) is changing fundamentally with disruptive implications for HPC storage. The classic modeling and simulation workloads running on supercomputers and large HPC clusters are joined by new workloads like machine learning and deep learning (ML/DL) running on mission- or business critical workflows on the same machine. In addition, high-performance data analysis (HPDA) of experimental and observational data from much more powerful instruments and machine-generated data from Internet of Things (IoT) sensors at the edge — soon connected by fifth-generation wireless (5G) cellular networks — is integrated into the workflow.

Cray calls this new era the exascale era and it’s characterized by converged workloads for crucial workflows running on the same machine, often simultaneously.

Cray saw this coming years ago and has developed the new Cray® ClusterStor® E1000 HPC/AI storage platform to enable HPC users to efficiently cope with the data explosion in this new paradigm.

Storage challenges facing today’s customers

Most customers have separate infrastructures for classic HPC (modeling and simulation workloads) and AI (ML/DL workloads) today.

  • Supercomputers and HPC clusters mainly based on compute nodes with CPUs served by HDD-based parallel HPC storage with storage capacity measured in petabytes (PB).
  • Artificial intelligence (AI) infrastructures mainly based on GPUs served by SSD-based enterprise network attached storage (NAS) with storage capacity measured in terabytes (TB).

The current status quo becomes a challenge once AI methods like ML/DL are running together with modeling and simulation workloads in workflows on supercomputers and HPC clusters. More than 50% of HPC users are already running machine learning programs today.[1] In general, HDD-based storage was designed for the write-intensive, sequential input/output (I/O) for large files. It is not well suited to serve the read-intensive I/O of a larger number of often small files that need to be accessed in random order during the training phases of ML/DL models.

As workloads converge, customers cannot easily stay on current HPC storage infrastructure. If they do, I/O bottlenecks will result in job pipeline congestion and expensive compute infrastructure will idle waiting for I/O to complete. At the same time, they also cannot scale their current AI storage — deployed in capacities of terabytes — to the petabyte scale capacities needed by supercomputer and HPC clusters running converged HPC and AI workloads. SSD storage is much more expensive than HDD storage today and IDC research shows that SSD in 2023 is projected to still be eight times more expensive than HDD in $/GB. [2]

Cray ClusterStor E1000 solves the exascale storage challenge

Customers should consider moving to the Cray® ClusterStor® E1000 HPC/AI storage solution.

“It is a flexible HPC/AI storage platform that leverages the strength of both SSD and HDD without incurring their respective weaknesses. ClusterStor E1000 solves the I/O bottleneck in the era of converged workloads without breaking the bank by orchestrating the data flow with the workflow through intelligent software. Cray saw the storage challenges of converged workloads coming as early as 2017 when development of the next generation ClusterStor storage system began,” says Uli Plechschmidt, director of storage product marketing for Cray.

Introducing the Cray ClusterStor E1000

Figure 1. Cray ClusterStor E1000 at a glance

Cray ClusterStor E1000 benefits

Performance: Meets the most demanding requirements of HPC/AI workloads with up to 1.6 TB/s and 50 million input/output operations per second (IOPS) per SSD rack and up to 120 GB/s and up to 10 PB usable capacity per HDD rack

Flexibility: All-SSD, all-HDD, and hybrid configurations orchestrating the data flow with the workflow through intelligent Cray software. Attaches to any supercomputer and any HPC cluster of any vendor via either 100/200 Gigabit Ethernet, EDR/HDR InfiniBand or the Cray Slingshot™ interconnect.

Scalability: Enables customers to start anywhere – as low as one-fifth of a rack – and to grow wherever they need to go. Up to and beyond the largest ClusterStor E1000 system already contracted (nearly 10 TB/s with more than 700 PB usable capacity).

Conclusion

The converged workflows of the exascale era break legacy HPC storage and AI storage models. The “cost of doing nothing” is either massive job pipeline congestion through I/O bottlenecks or spending most of the total system budget on storage. A storage solution is needed for the new era.

“Unlike HPC/AI storage-only vendors we believe that storage is a means to an end. The end being insights and breakthroughs that can make the world a better place to work and live in,” says Plechschmidt. “Cray has architected the ClusterStor E1000 in a way that can meet the storage requirements of converged workloads in the most efficient way. Exploiting the strengths of HDD and SSD without incurring their weaknesses — for HDD the challenge to serve random I/O of large numbers of small files and for SSD the high cost per gigabyte — by intelligent hardware design and through intelligent software that aligns the data flow with the workflow. Due to that differentiated design, customers who choose the ClusterStor E1000 typically will be able to achieve their storage requirements in a more efficient way — enabling them to spend more of their total system budget on CPU/GPU compute nodes when compared to alternative HPC/AI storage offerings.”

—————————————————————————————————————————–

References

1] Intersect360 HPC User Budget Map Survey: Machine Learning’s Impact on HPC Environments, September 2018.

2] IDC, Worldwide 2019-2023 Enterprise SSD and HDD Combined Market Overview, June 2019.

 

About Cray

About Cray Inc.

Cray, a Hewlett Packard Enterprise company, combines computation and creativity so visionaries can keep asking questions that challenge the limits of possibility. Drawing on more than 45 years of experience, Cray develops the world’s most advanced supercomputers, pushing the boundaries of performance, efficiency and scalability. Cray continues to innovate today at the convergence of data and discovery, offering a comprehensive portfolio of supercomputers, high-performance storage, data analytics and artificial intelligence solutions.  Go to www.cray.com for more information.

CRAY and ClusterStor are registered trademarks of Cray Inc. in the United States and other countries, and Shasta and Slingshot are trademarks of Cray Inc. Other product and service names mentioned herein are the trademarks of their respective owners.

Copyright 2019

 

Subscribe to HPCwire's Weekly Update!

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

Graphcore Introduces Larger-Than-Ever IPU-Based Pods

October 22, 2021

Graphcore and its “Intelligent Processing Units” (IPUs) emerged from stealth in 2016 and launched its second-generation IPU in 2020. While the company has also launched its IPUs in a variety of form factors over the Read more…

Quantum Chemistry Project to Be Among the First on EuroHPC’s LUMI System

October 22, 2021

Finland’s CSC has just installed the first module of LUMI, a 550-peak petaflops system supported by the European Union’s EuroHPC Joint Undertaking. While LUMI -- pictured in the header -- isn’t slated to complete i Read more…

Killer Instinct: AMD’s Multi-Chip MI200 GPU Readies for a Major Global Debut

October 21, 2021

AMD’s next-generation supercomputer GPU is on its way – and by all appearances, it’s about to make a name for itself. The AMD Radeon Instinct MI200 GPU (a successor to the MI100) will, over the next year, begin to power three massive systems on three continents: the United States’ exascale Frontier system; the European Union’s pre-exascale LUMI system; and Australia’s petascale Setonix system. Read more…

D-Wave Embraces Gate-Based Quantum Computing; Charts Path Forward

October 21, 2021

Earlier this month D-Wave Systems, the quantum computing pioneer that has long championed quantum annealing-based quantum computing (and sometimes taken heat for that approach), announced it was expanding into gate-based Read more…

LLNL Prepares the Water and Power Infrastructure for El Capitan

October 21, 2021

When it’s (ostensibly) ready in early 2023, El Capitan is expected to deliver in excess of two exaflops of peak computing power – around four times the power of Fugaku, the current top-ranked supercomputer in the wor Read more…

AWS Solution Channel

Royalty-free stock illustration ID: 537899029

Running GROMACS on GPU instances

Comparing the performance of real applications across different Amazon Elastic Compute Cloud (Amazon EC2) instance types is the best way we’ve found for finding optimal configurations for HPC applications here at AWS. Read more…

Faster Optical Switch that Operates at ‘Room Temp’ Developed by IBM, Skolkovo Researchers

October 19, 2021

Optical switching technology holds great promise for many applications but hot operating temperatures have been a key obstacle slowing progress. Now, a new optical switching device that can operate at room temperatures a Read more…

Killer Instinct: AMD’s Multi-Chip MI200 GPU Readies for a Major Global Debut

October 21, 2021

AMD’s next-generation supercomputer GPU is on its way – and by all appearances, it’s about to make a name for itself. The AMD Radeon Instinct MI200 GPU (a successor to the MI100) will, over the next year, begin to power three massive systems on three continents: the United States’ exascale Frontier system; the European Union’s pre-exascale LUMI system; and Australia’s petascale Setonix system. Read more…

D-Wave Embraces Gate-Based Quantum Computing; Charts Path Forward

October 21, 2021

Earlier this month D-Wave Systems, the quantum computing pioneer that has long championed quantum annealing-based quantum computing (and sometimes taken heat fo Read more…

LLNL Prepares the Water and Power Infrastructure for El Capitan

October 21, 2021

When it’s (ostensibly) ready in early 2023, El Capitan is expected to deliver in excess of two exaflops of peak computing power – around four times the powe Read more…

Intel Reorgs HPC Group, Creates Two ‘Super Compute’ Groups

October 15, 2021

Following on changes made in June that moved Intel’s HPC unit out of the Data Platform Group and into the newly created Accelerated Computing Systems and Graphics (AXG) business unit, led by Raja Koduri, Intel is making further updates to the HPC group and announcing... Read more…

Quantum Workforce – NSTC Report Highlights Need for International Talent

October 13, 2021

Attracting and training the needed quantum workforce to fuel the ongoing quantum information sciences (QIS) revolution is a hot topic these days. Last week, the U.S. National Science and Technology Council issued a report – The Role of International Talent in Quantum Information Science... Read more…

Eni Returns to HPE for ‘HPC4’ Refresh via GreenLake

October 13, 2021

Italian energy company Eni is upgrading its HPC4 system with new gear from HPE that will be installed in Eni’s Green Data Center in Ferrera Erbognone (a provi Read more…

The Blueprint for the National Strategic Computing Reserve

October 12, 2021

Over the last year, the HPC community has been buzzing with the possibility of a National Strategic Computing Reserve (NSCR). An in-utero brainchild of the COVID-19 High-Performance Computing Consortium, an NSCR would serve as a Merchant Marine for urgent computing... Read more…

UCLA Researchers Report Largest Chiplet Design and Early Prototyping

October 12, 2021

What’s the best path forward for large-scale chip/system integration? Good question. Cerebras has set a high bar with its wafer scale engine 2 (WSE-2); it has 2.6 trillion transistors, including 850,000 cores, and was fabricated using TSMC’s 7nm process on a roughly 8” x 8” silicon footprint. Read more…

Enter Dojo: Tesla Reveals Design for Modular Supercomputer & D1 Chip

August 20, 2021

Two months ago, Tesla revealed a massive GPU cluster that it said was “roughly the number five supercomputer in the world,” and which was just a precursor to Tesla’s real supercomputing moonshot: the long-rumored, little-detailed Dojo system. Read more…

Esperanto, Silicon in Hand, Champions the Efficiency of Its 1,092-Core RISC-V Chip

August 27, 2021

Esperanto Technologies made waves last December when it announced ET-SoC-1, a new RISC-V-based chip aimed at machine learning that packed nearly 1,100 cores onto a package small enough to fit six times over on a single PCIe card. Now, Esperanto is back, silicon in-hand and taking aim... Read more…

US Closes in on Exascale: Frontier Installation Is Underway

September 29, 2021

At the Advanced Scientific Computing Advisory Committee (ASCAC) meeting, held by Zoom this week (Sept. 29-30), it was revealed that the Frontier supercomputer is currently being installed at Oak Ridge National Laboratory in Oak Ridge, Tenn. The staff at the Oak Ridge Leadership... Read more…

Ahead of ‘Dojo,’ Tesla Reveals Its Massive Precursor Supercomputer

June 22, 2021

In spring 2019, Tesla made cryptic reference to a project called Dojo, a “super-powerful training computer” for video data processing. Then, in summer 2020, Tesla CEO Elon Musk tweeted: “Tesla is developing a [neural network] training computer... Read more…

Intel Reorgs HPC Group, Creates Two ‘Super Compute’ Groups

October 15, 2021

Following on changes made in June that moved Intel’s HPC unit out of the Data Platform Group and into the newly created Accelerated Computing Systems and Graphics (AXG) business unit, led by Raja Koduri, Intel is making further updates to the HPC group and announcing... Read more…

Intel Completes LLVM Adoption; Will End Updates to Classic C/C++ Compilers in Future

August 10, 2021

Intel reported in a blog this week that its adoption of the open source LLVM architecture for Intel’s C/C++ compiler is complete. The transition is part of In Read more…

Hot Chips: Here Come the DPUs and IPUs from Arm, Nvidia and Intel

August 25, 2021

The emergence of data processing units (DPU) and infrastructure processing units (IPU) as potentially important pieces in cloud and datacenter architectures was Read more…

AMD-Xilinx Deal Gains UK, EU Approvals — China’s Decision Still Pending

July 1, 2021

AMD’s planned acquisition of FPGA maker Xilinx is now in the hands of Chinese regulators after needed antitrust approvals for the $35 billion deal were receiv Read more…

Leading Solution Providers

Contributors

HPE Wins $2B GreenLake HPC-as-a-Service Deal with NSA

September 1, 2021

In the heated, oft-contentious, government IT space, HPE has won a massive $2 billion contract to provide HPC and AI services to the United States’ National Security Agency (NSA). Following on the heels of the now-canceled $10 billion JEDI contract (reissued as JWCC) and a $10 billion... Read more…

10nm, 7nm, 5nm…. Should the Chip Nanometer Metric Be Replaced?

June 1, 2020

The biggest cool factor in server chips is the nanometer. AMD beating Intel to a CPU built on a 7nm process node* – with 5nm and 3nm on the way – has been i Read more…

Quantum Roundup: IBM, Rigetti, Phasecraft, Oxford QC, China, and More

July 13, 2021

IBM yesterday announced a proof for a quantum ML algorithm. A week ago, it unveiled a new topology for its quantum processors. Last Friday, the Technical Univer Read more…

The Latest MLPerf Inference Results: Nvidia GPUs Hold Sway but Here Come CPUs and Intel

September 22, 2021

The latest round of MLPerf inference benchmark (v 1.1) results was released today and Nvidia again dominated, sweeping the top spots in the closed (apples-to-ap Read more…

Julia Update: Adoption Keeps Climbing; Is It a Python Challenger?

January 13, 2021

The rapid adoption of Julia, the open source, high level programing language with roots at MIT, shows no sign of slowing according to data from Julialang.org. I Read more…

Intel Unveils New Node Names; Sapphire Rapids Is Now an ‘Intel 7’ CPU

July 27, 2021

What's a preeminent chip company to do when its process node technology lags the competition by (roughly) one generation, but outmoded naming conventions make i Read more…

Frontier to Meet 20MW Exascale Power Target Set by DARPA in 2008

July 14, 2021

After more than a decade of planning, the United States’ first exascale computer, Frontier, is set to arrive at Oak Ridge National Laboratory (ORNL) later this year. Crossing this “1,000x” horizon required overcoming four major challenges: power demand, reliability, extreme parallelism and data movement. Read more…

Quantum Computer Market Headed to $830M in 2024

September 13, 2021

What is one to make of the quantum computing market? Energized (lots of funding) but still chaotic and advancing in unpredictable ways (e.g. competing qubit tec Read more…

  • arrow
  • Click Here for More Headlines
  • arrow
HPCwire