Cray Debuts ClusterStor E1000 Finishing Remake of Portfolio for ‘Exascale Era’

By John Russell

October 30, 2019

Cray, now owned by HPE, today introduced the ClusterStor E1000 storage platform, which leverages Cray software and mixes hard disk drives (HDD) and flash memory (SSD) to accommodate converged HPC-AI workloads at many scales. Cray described the new system as the final step in re-architecting its product portfolio for the exascale era. Cray also didn’t miss the opportunity to tout selection of ClusterStor E1000 for use in the first three planned U.S. exascale computers (Aurora, Frontier, and El Capitan with Cray as the prime systems contractor for the latter two).

For a company whose long-term health was sometimes questioned in recent years, Cray is riding high with prior worries about its standalone viability decisively resolved by the HPE acquisition. In a press pre-briefing, Uli Plechschmidt, director of storage product marketing, said the ClusterStor E1000 completes Cray’s product refresh – joining Cray’s Shasta architecture, Slingshot interconnect technology, and updated Cray software. He also reinforced Cray’s intent to serve a larger market.

“The exascale era is not defined by the size of the systems. It’s defined by the convergence of workloads – not just classic modeling and simulation running on the supercomputers or HPC clusters, but also methods of artificial intelligence like machine learning or big data analytics, all running on one machine. Going forward we’re going to focus on artificial intelligence joining the modeling and simulation because it’s the most disruptive workload from an I/O pattern perspective,” he said.

Cray expects new dense media packaging, a newly-designed memory controller, and new software (ClusterStor Data Services) – all tightly integrated with its system architecture – to help tame the data-flow challenge. Interestingly, the new controller uses AMD’s newest Rome generation CPU.

“To handle the massive growth in data that corporations worldwide are dealing with in their digital transformations, a completely new approach to storage is required,” said Peter Ungaro, SVP & GM, HPC & AI of Cray, a Hewlett Packard Enterprise company. “Cray’s new storage platform is a comprehensive rethinking of what high performance storage means for the Exascale Era. The intelligent software and hardware design of ClusterStor E1000 orchestrates the data flow with the workflow – that’s something no other solution on the market can do.”

The ClusterStor E1000 is the latest addition to the ClusterStor product line, which Cray purchased from Seagate in 2017. Claiming the new system is the fastest in the world on several metrics, Plechschmidt said new dense storage media packaging could deliver up to 1.6TB/s and 50 million IOPS per solid state drive rack and up to 120 GB/s and 10 PB usable per hard disk drive base rack.

Also noteworthy, NERSC (National Energy Research Scientific Computing Center) will deploy the new ClusterStor E1000 on Perlmutter as its fast all flash storage tier, which will be capable of over four terabytes per second write bandwidth. “This architecture will support our diverse workloads and research disciplines,” said NERSC Director Sudip Dosanjh. “Because this file system will be the first all-NVMe file system deployed at a scale of 30 petabytes usable capacity, extensive quantitative analysis was undertaken by NERSC to determine the optimal architecture to support the workflows our researchers and scientists use across biology, environment, chemistry, nuclear physics, fusion energy, plasma physics and computing research.”

The rise of mixed workloads to support converged HPC/AI workloads is especially challenging for the storage system. As a general rule the sequential movement of large files typical of traditional modeling and simulation is reasonably well-served by less expensive hard disk drives. Conversely, the I/O management for AI (particularly training) often requires moving large numbers of small files, accessed at random, which is best handled by more expensive flash memory.

“IDC projects that even in the year 2023 there still will be an 8X price difference per gigabyte between hard disk drives and enterprise SSD. [Although] we actually think the SSD price performance improvement will go a little bit faster,” said Plechschmidt. Balancing cost-performance issue is a moving target with data management software as important as the hardware media.

Broadly, ClusterStor E1000 is a “factory engineered solution” in which users mix and match media to accommodate the specific I/O profiles of the workloads in the workflow. It can be configured three ways (list of scalable storage units (SSU) at the end of the article):

  • All Flash configuration. SSU-Fs only with up to 1.6 TB/sec (sequential read) and up to 50 million IOPS per storage rack.
  • All Disk configuration.
 SSU-Ds only with up to 120 GB/sec and up to 10 PB usable capacity per storage rack
  • Hybrid configuration.
 SSU-Fs as Flash pool and SSU-Ds as HDD pool in the same file system with ClusterStor Data Services software to orchestrate the data flow with the workflow in the most efficient way.

“The base rack contains at a minimum one 2U System Management Unit and one 2U Metadata Unit. To scale out, SSUs are added — at first filing up the remaining space in the base rack and then populating expansion racks,” according to Plechschmidt.

Here’s bit more technical detail from Cray literature:

“The new high-density enclosure will support 106 large form factor HDDs in 4 RU. Initially, the enclosure will support 4, 6, 10, 12, 14 and 16 TB HDDs followed by 20 TB and 24 TB HAMR drives respectively in the 2021 timeframe. The 2U storage controllers are using a single HDR/Slingshot NIC and 2x PCIe gen 4 SAS HBAs (16 lanes of 12 Gbit SAS) per node connecting to HDD enclosures. Initially, storage controller will support 1, 2 or 4 HDD enclosures. The HDD enclosure is designed as a fully redundant storage array with 2x SAS adapters and 2x fully configured side planes.

“As with current ClusterStor solutions, each HDD enclosure will support two disk pools (consisting of 53 drives each) configured with either GridRAID (when using LDISKFS) or ZFS…The main concept when configuring a ClusterStor E1000 solution is to combine flash and HDD based Scalable Storage Units to achieve the right balance between capacity and performance. The idea is to use Flash-based Scalable Storage Units (SSU-F) to achieve the peak performance and HDD-based Scalable Storage Units (D-SSU-x) composing the required usable storage volume. By combining these two volumes into a single namespace and using the intelligent software ClusterStor Data Services there no longer will be a need to overprovision a system to achieve performance.”

Cray’s new storage controller and software, ClusterStor Data Services, enable users to align the data flow with their specific workflows; in essence place the application data at the right time on the right storage media (SSD pool or HDD pool) in the file system.

Plechschmidt said, “ClusterStor Data services enables customers to orchestrate the data flow within the file system between the flex pool. There are three modes and we’re rolling them out in a phased approach”:

  • Scripted Tiering. Available at launch, allows power users to have total control (command line interface) over what data gets staged to what storage media and when. Cray is providing tools to help manage the process.
  • Scheduled Tiering.Expected in mid-2020, this is an automated method to promote data to flash then demote it back to HDD. “You can do it in the workload manager – a user needs to insert a couple of very simple workload manager directives and then, completely outside of the job runtime and application runtime, data gets promoted up, the job runs, and once the job completed, outside of the application runtime, the data gets demoted back to make space now for the next job.”
  • Transparent Tiering. On the roadmap for the first half of 2021, transparent tiering won’t require user input. It’s completely transparent and to the user, said Plechschmidt, although presumably some level of control will be possible.

For connectivity, it’s no surprise that ClusterStor E1000 has 200 Gbps Slingshot switches on top of the rack. Cray also provides 100/200 Gbps Ethernet and InfiniBand EDR/HDR connectivity. Lustre is the embedded file system.

“We have chosen to continue to embed Lustre because it simply is the highest-performance lowest-cost parallel file system for HPC and we’re hardening that for reliability at exascale,” said Plechschmidt, emphasizing the new platform is flexible and able to “embed other file systems or emerging file systems in the future [but] we are convinced that Lustre is the right file system for the foreseeable future.”

Entry configurations of Cray ClusterStor E1000 will start at around $200,000 including three years of basic customer support. “The first early access systems are planned to ship to customers in December 2019. First customer shipment for production systems is planned for Q1 CY2020 with full production availability planned for Q2 calendar year 2020,” said Plechschmidt.

 

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!

NIST/Xanadu Researchers Report Photonic Quantum Computing Advance

March 3, 2021

Researchers from the National Institute of Standards and Technology (NIST) and Xanadu, a young Canada-based quantum computing company, have reported developing a full-stack, photonic quantum computer able to carry out th Read more…

By John Russell

Can Deep Learning Replace Numerical Weather Prediction?

March 3, 2021

Numerical weather prediction (NWP) is a mainstay of supercomputing. Some of the first applications of the first supercomputers dealt with climate modeling, and even to this day, the largest climate models are heavily con Read more…

By Oliver Peckham

HPE Names Justin Hotard New HPC Chief as Pete Ungaro Departs

March 2, 2021

HPE CEO Antonio Neri announced today (March 2, 2020) the appointment of Justin Hotard as general manager of HPC, mission critical solutions and labs, effective immediately. Hotard replaces long-time Cray exec Pete Ungaro Read more…

By Tiffany Trader

ORNL’s Jeffrey Vetter on How IRIS Runtime will Help Deal with Extreme Heterogeneity

March 2, 2021

Jeffery Vetter is a familiar figure in HPC. Last year he became one of the new section heads in a reorganization at Oak Ridge National Laboratory. He had been founding director of ORNL's Future Technologies Group which i Read more…

By John Russell

HPC Career Notes: March 2021 Edition

March 1, 2021

In this monthly feature, we’ll keep you up-to-date on the latest career developments for individuals in the high-performance computing community. Whether it’s a promotion, new company hire, or even an accolade, we’ Read more…

By Mariana Iriarte

AWS Solution Channel

Moderna Accelerates COVID-19 Vaccine Development on AWS

Marcello Damiani, Chief Digital and Operational Excellence Officer at Moderna, joins Todd Weatherby, Vice President of AWS Professional Services Worldwide, for a discussion on developing Moderna’s COVID-19 vaccine, scaling systems to enable global distribution, and leveraging cloud technologies to accelerate processes. Read more…

Supercomputers Enable First Holistic Model of SARS-CoV-2, Showing Spike Proteins Move in Tandem

February 28, 2021

Most models of SARS-CoV-2, the coronavirus that causes COVID-19, hone in on key features of the virus: for instance, the spike protein. Some of this is attributable to the relative importance of those features, but most Read more…

By Oliver Peckham

Can Deep Learning Replace Numerical Weather Prediction?

March 3, 2021

Numerical weather prediction (NWP) is a mainstay of supercomputing. Some of the first applications of the first supercomputers dealt with climate modeling, and Read more…

By Oliver Peckham

HPE Names Justin Hotard New HPC Chief as Pete Ungaro Departs

March 2, 2021

HPE CEO Antonio Neri announced today (March 2, 2020) the appointment of Justin Hotard as general manager of HPC, mission critical solutions and labs, effective Read more…

By Tiffany Trader

ORNL’s Jeffrey Vetter on How IRIS Runtime will Help Deal with Extreme Heterogeneity

March 2, 2021

Jeffery Vetter is a familiar figure in HPC. Last year he became one of the new section heads in a reorganization at Oak Ridge National Laboratory. He had been f Read more…

By John Russell

HPC Career Notes: March 2021 Edition

March 1, 2021

In this monthly feature, we’ll keep you up-to-date on the latest career developments for individuals in the high-performance computing community. Whether it Read more…

By Mariana Iriarte

African Supercomputing Center Inaugurates ‘Toubkal,’ Most Powerful Supercomputer on the Continent

February 25, 2021

Historically, Africa hasn’t exactly been synonymous with supercomputing. There are only a handful of supercomputers on the continent, with few ranking on the Read more…

By Oliver Peckham

Japan to Debut Integrated Fujitsu HPC/AI Supercomputer This Spring

February 25, 2021

The integrated Fujitsu HPC/AI Supercomputer, Wisteria, is coming to Japan this spring. The University of Tokyo is preparing to deploy a heterogeneous computing Read more…

By Tiffany Trader

Xilinx Launches Alveo SN1000 SmartNIC

February 24, 2021

FPGA vendor Xilinx has debuted its latest SmartNIC model, the Alveo SN1000, with integrated “composability” features that allow enterprise users to add their own custom networking functions to supplement its built-in networking. By providing deep flexibility... Read more…

By Todd R. Weiss

ASF Keynotes Showcase How HPC and Big Data Have Pervaded the Pandemic

February 24, 2021

Last Thursday, a range of experts joined the Advanced Scale Forum (ASF) in a rapid-fire roundtable to discuss how advanced technologies have transformed the way humanity responded to the COVID-19 pandemic in indelible ways. The roundtable, held near the one-year mark of the first... Read more…

By Oliver Peckham

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…

By John Russell

Esperanto Unveils ML Chip with Nearly 1,100 RISC-V Cores

December 8, 2020

At the RISC-V Summit today, Art Swift, CEO of Esperanto Technologies, announced a new, RISC-V based chip aimed at machine learning and containing nearly 1,100 low-power cores based on the open-source RISC-V architecture. Esperanto Technologies, headquartered in... Read more…

By Oliver Peckham

Azure Scaled to Record 86,400 Cores for Molecular Dynamics

November 20, 2020

A new record for HPC scaling on the public cloud has been achieved on Microsoft Azure. Led by Dr. Jer-Ming Chia, the cloud provider partnered with the Beckman I Read more…

By Oliver Peckham

Programming the Soon-to-Be World’s Fastest Supercomputer, Frontier

January 5, 2021

What’s it like designing an app for the world’s fastest supercomputer, set to come online in the United States in 2021? The University of Delaware’s Sunita Chandrasekaran is leading an elite international team in just that task. Chandrasekaran, assistant professor of computer and information sciences, recently was named... Read more…

By Tracey Bryant

NICS Unleashes ‘Kraken’ Supercomputer

April 4, 2008

A Cray XT4 supercomputer, dubbed Kraken, is scheduled to come online in mid-summer at the National Institute for Computational Sciences (NICS). The soon-to-be petascale system, and the resulting NICS organization, are the result of an NSF Track II award of $65 million to the University of Tennessee and its partners to provide next-generation supercomputing for the nation's science community. 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…

By Doug Black

Top500: Fugaku Keeps Crown, Nvidia’s Selene Climbs to #5

November 16, 2020

With the publication of the 56th Top500 list today from SC20's virtual proceedings, Japan's Fugaku supercomputer – now fully deployed – notches another win, Read more…

By Tiffany Trader

Gordon Bell Special Prize Goes to Massive SARS-CoV-2 Simulations

November 19, 2020

2020 has proven a harrowing year – but it has produced remarkable heroes. To that end, this year, the Association for Computing Machinery (ACM) introduced the Read more…

By Oliver Peckham

Leading Solution Providers

Contributors

Texas A&M Announces Flagship ‘Grace’ Supercomputer

November 9, 2020

Texas A&M University has announced its next flagship system: Grace. The new supercomputer, named for legendary programming pioneer Grace Hopper, is replacing the Ada system (itself named for mathematician Ada Lovelace) as the primary workhorse for Texas A&M’s High Performance Research Computing (HPRC). Read more…

By Oliver Peckham

Saudi Aramco Unveils Dammam 7, Its New Top Ten Supercomputer

January 21, 2021

By revenue, oil and gas giant Saudi Aramco is one of the largest companies in the world, and it has historically employed commensurate amounts of supercomputing Read more…

By Oliver Peckham

Intel Xe-HP GPU Deployed for Aurora Exascale Development

November 17, 2020

At SC20, Intel announced that it is making its Xe-HP high performance discrete GPUs available to early access developers. Notably, the new chips have been deplo Read more…

By Tiffany Trader

Intel Teases Ice Lake-SP, Shows Competitive Benchmarking

November 17, 2020

At SC20 this week, Intel teased its forthcoming third-generation Xeon "Ice Lake-SP" server processor, claiming competitive benchmarking results against AMD's second-generation Epyc "Rome" processor. Ice Lake-SP, Intel's first server processor with 10nm technology... Read more…

By Tiffany Trader

New Deep Learning Algorithm Solves Rubik’s Cube

July 25, 2018

Solving (and attempting to solve) Rubik’s Cube has delighted millions of puzzle lovers since 1974 when the cube was invented by Hungarian sculptor and archite Read more…

By John Russell

Livermore’s El Capitan Supercomputer to Debut HPE ‘Rabbit’ Near Node Local Storage

February 18, 2021

A near node local storage innovation called Rabbit factored heavily into Lawrence Livermore National Laboratory’s decision to select Cray’s proposal for its CORAL-2 machine, the lab’s first exascale-class supercomputer, El Capitan. Details of this new storage technology were revealed... Read more…

By Tiffany Trader

It’s Fugaku vs. COVID-19: How the World’s Top Supercomputer Is Shaping Our New Normal

November 9, 2020

Fugaku is currently the most powerful publicly ranked supercomputer in the world – but we weren’t supposed to have it yet. The supercomputer, situated at Japan’s Riken scientific research institute, was scheduled to come online in 2021. When the pandemic struck... Read more…

By Oliver Peckham

African Supercomputing Center Inaugurates ‘Toubkal,’ Most Powerful Supercomputer on the Continent

February 25, 2021

Historically, Africa hasn’t exactly been synonymous with supercomputing. There are only a handful of supercomputers on the continent, with few ranking on the Read more…

By Oliver Peckham

  • arrow
  • Click Here for More Headlines
  • arrow
HPCwire