Infortrend optimizes data center footprint with high density 4U 60-bay storage

Infortrend Technology Inc. has introduced the 4U 60-bay unified storage system to address rapid data growth and limited datacenter space challenges. The 4U 60-bay EonStor GS Gen2 features advanced performance, and its high density enclosure can accommodate up to 896 drives in one system, delivering 10PB+ massive capacity for data-intensive applications.

EonStor GS 3000/4000 Gen2 is the latest generation of Infortrend’s high availability enterprise unified storage which integrates NAS/SAN/Cloud gateway into one system. 

Compared with the previous solutions, GS Gen2 NAS performance is improved by 60 percent, reaching up to 9,900/5,500 MB/s read/write throughput. It also supports 900K IOPS performance. Other enhancements include double memory capacity, four onboard 10GbE ports for high-speed I/O connections, and comprehensive form factors.

The EonStor GS 3000/2000 series are comprehensive enterprise solutions with exceptional SAN and NAS transmission performance to meet the storage needs of SMBs to large enterprises. Besides various form factors, they also offer up to 4 sets of modular host boards to fulfill the requirements of numerous applications or architecture transformation to allow more deployment flexibility. 

Furthermore, the system also comes with high expandability of more than 8PB storage capacity to easily meet the needs of future data growth. As such, its outstanding performance, flexibility, and high expandability can run critical applications for SAN and NAS on the market. The GS 3000 is perfect for the performance-conscious, while the GS 2000 is ideal for the budget—conscious as it can meet all general storage needs.

With the support of Intel’s advanced CPU platform, and the newly-released 4U 60-bay form factor, GS Gen2 is designed to empower customers to overcome the issues of rapid data volume growth in terms of storage performance and capacity, while optimizing datacenter space utilization.   

The EonStor GS 5000/4000 series forms the backbone of our high performance solution for enterprises with its powerful performance, flexibility, and high expandability to efficiently boost overall productivity. They can handle large amounts of I/O and file transfers even under high workloads, and are especially ideal for hybrid environments that adopt SAN, NAS, and cloud integration. 

The GS 5000 is perfect for those who want performance, capacity, and a high port count, while the GS 4000 is ideal for the budget-conscious as it can easily meet all general storage needs.

“Infortrend releases the 4U 60-bay GS Gen2 series to prepare businesses for the enormous data management requirements,” said Thomas Kao, senior director of product planning. “This new solution is especially future-proof for data-intensive applications, such as media & entertainment, surveillance, and HPC (high performance computing).”

NVIDIA releases Magnum IO software line to help reduce data bottlenecks for data scientists and AI, HPC researchers

NVIDIA introduced NVIDIA Magnum IO, a suite of software to help data scientists and AI and high performance computing researchers process massive amounts of data in minutes, rather than hours. This offering ahs been developed in close collaboration with vendors in networking and storage, including DataDirect Networks, Excelero, IBM, Mellanox and WekaIO.

Optimized to eliminate storage and input/output bottlenecks, Magnum IO delivers up to 20 times faster data processing for multi-server, multi-GPU computing nodes when working with massive datasets to carry out complex financial analysis, climate modeling and other HPC workloads.

At the heart of Magnum IO is GPUDirect, which provides a path for data to bypass CPUs and travel on “open highways” offered by GPUs, storage and networking devices. Compatible with a wide range of communications interconnects and APIs — including NVIDIA NVLink and NCCL, as well as OpenMPI and UCX — GPUDirect is composed of peer-to-peer and RDMA elements.

Its newest element is GPUDirect Storage, which enables researchers to bypass CPUs when accessing storage and quickly access data files for simulation, analysis or visualization.

Magnum IO integrates infrastructure elements to maximize storage and network I/O performance and functionality. It optimizes IO performance by bypassing the CPU to enable direct IO between GPU memory and network storage. It also relieves CPU contention to create a more balanced GPU-accelerated system and delivers peak IO bandwidth with up to ten times fewer CPU cores, and provides optimized implementation for current and future platforms, whether the data transfers are latency sensitive, bandwidth sensitive, or collectives.

NVIDIA Magnum IO software is available, with the exception of GPUDirect Storage, which is currently available to select early-access customers. Broader release of GPUDirect Storage is planned for the first half of next year.

HPE, Cray release comprehensive HPC and AI offerings optimized for exascale computing on premises, in the cloud, or as-a-service

Hewlett Packard Enterprise (HPE) announced Thursday that it will deliver comprehensive high-performance computing (HPC) and Artificial Intelligence (AI) portfolio for the exascale era, which is characterized by explosive data growth and new converged workloads such as HPC, AI, and analytics.

The addition of Cray Inc., which HPE recently acquired, bolsters HPE’s HPC and AI solutions to now encompass an end-to-end supercomputing architecture across compute, interconnect, software, storage and services, delivered on premises, hybrid or as-a-Service. Now every enterprise can leverage the same foundational HPC technologies that power the world’s fastest systems, and integrate them into their data centers to unlock insights and fuel new discovery.

Digital transformation is driving new data-intensive workloads and real-time analytics operating at an unprecedented scale. New software, compute, interconnect, and storage capabilities are required for customers to unlock the potential of their data and accelerate innovation. 

HPE delivers solutions for any experience from single, small systems all the way to exascale-class supercomputers with tailored software, interconnect and storage capabilities. This includes solutions for modeling and simulation in weather forecasting, manufacturing and energy sectors, and AI and big data analytics in precision medicine, autonomous vehicles, geospatial imaging and financial services.

Rescale debuts ScaleX Government, its federally compliant HPC cloud service

Rescale announced Wednesday that they are the initial company to deliver a completely cloud-based HPC platform for the public sector. Rescale’s latest offering, ScaleX Government, is the only fully managed HPC service that delivers the scalability of the cloud with FedRAMP Moderate (In-Process), ITAR compliance, or both.

With this announcement, Rescale can now be found in the official FedRAMP Marketplace. This enables federal engineers, scientists and researchers to shift outdated, fixed IT infrastructure to the latest in on-demand cloud architectures and meet their compliance requirements. 

The National Renewable Energy Lab (NREL), a division of the U.S. Department of Energy (DOE), co-sponsored the FedRAMP application with Rescale, to accelerate its energy efficiency and exploration research. HPC in the cloud makes it possible to instantly scale workloads up or down as needed, without requiring resource management that often results in long wait times. This ultimately leads to better research, faster results, and more efficient spending.

Until now, HPC in the cloud has been inaccessible for all Federal Agencies and many research institutions because cloud products and services cannot be adopted without FedRAMP status. ScaleX Government delivers the first cloud HPC service that is compliant end-to-end across every constituent element, including compute, storage, application management, and orchestration stack.

Intel debuts GPU architecture with HPC and AI acceleration, and oneAPI software stack for heterogeneous architectures

Intel unveiled its vision for extending its leadership in the convergence of high-performance computing (HPC) and artificial intelligence (AI) with additions to its data-centric silicon portfolio and an ambitious new software initiative that represents a paradigm shift from single-architecture, single-vendor programming models.

Addressing the increasing use of heterogeneous architectures in high-performance computing, Intel expanded on its existing technology portfolio to move, store and process data more effectively by announcing a new category of discrete general-purpose GPUs optimized for AI and HPC convergence. 

Intel also launched the oneAPI industry initiative to deliver a unified and simplified programming model for application development across heterogenous processing architectures, including CPUs, GPUs, FPGAs and other accelerators. The launch of oneAPI represents millions of Intel engineering hours in software development and marks a game-changing evolution from limiting, proprietary programming approaches to an open standards-based model for cross-architecture developer engagement and innovation.

The oneAPI initiative Intel launched will define programming for an increasingly AI-infused, multi-architecture world. oneAPI delivers a unified and open programming experience to developers on the architecture of their choice without compromising performance and eliminating the complexity of separate code bases, multiple-programming languages, and different tools and workflows. oneAPI preserves existing software investments with support for existing languages, while delivering flexibility for developers to create versatile applications.

oneAPI includes both an industry initiative based on open specifications and an Intel beta product. The oneAPI specification includes a direct programming language, APIs and a low-level hardware interface. Intel’s oneAPI beta software provides developers a comprehensive portfolio of developer tools that include compilers, libraries and analyzers, packaged into domain-focused toolkits. 

The initial oneAPI beta release targets Intel Xeon scalable processors, Intel Core processors with integrated graphics, and Intel FPGAs, with additional hardware support to follow in future releases.

nCorium, KIOXIA America, EchoStreams come together to deliver high-performance storage platform based on PCIe 4.0 NVMe SSDs

nCorium, in collaboration with KIOXIA America, Inc.(formerly Toshiba Memory America, Inc.), the U.S.-based subsidiary of KIOXIA Corporation, and EchoStreams Innovative Solutions Inc. announced a new storage acceleration platform. 

Based on nCorium’s second-generation, near-memory acceleration engines, KIOXIA’s PCIe 4.0 NVMe SSDs and EchoStreams’ FlacheStreams Storage Server, this fully integrated solution is designed to perform storage functions in-line and stripe data to local or remote flash storage.

Applications such as high-performance computing (HPC), artificial intelligence, deep learning, data analytics, computer-aided engineering, climate modeling, and entertainment rely on parallel file systems to achieve high-throughput storage they require. 

As disk drives in parallel file systems are rapidly replaced by high-bandwidth SSDs, the processing capabilities of traditional architectures simply fail to keep up. nCorium’s storage acceleration platform delivers improved capabilities by offloading compute-intensive functions to the near memory acceleration engines called nCorium Compute in Memory Modules (nCIMMs).

Mellanox Skyway 200 Gigabit HDR InfiniBand to Ethernet gateway appliance released for high performance and cloud data centers

Mellanox Technologies introduced Mellanox Skyway, a 200 gigabit HDR InfiniBand to Ethernet gateway appliance. Mellanox Skyway enables a scalable and efficient way to connect the high-performance, low-latency InfiniBand data center to external Ethernet infrastructures or connectivity.

Mellanox Skyway is the next generation of the existing 56 gigabit FDR InfiniBand to 40 gigabit Ethernet gateway system, deployed in multiple data centers around the world.

High performance computing, artificial intelligence, data-intensive and cloud infrastructures leverage InfiniBand's high data throughput, extremely low latency, and smart in-network computing acceleration engines to deliver critical application performance and scalability. 

For cases requiring connectivity into external Ethernet-based platforms, Mellanox offers the Skyway appliance, a 1.6 terabit-per-second InfiniBand to Ethernet gateway appliance, supporting eight 100 or 200 gigabit ports on each of the InfiniBand and Ethernet sides. IT managers can scale the number of Skyway appliances over time to meet the demands of their users.

By leveraging the Mellanox Skyway appliance, InfiniBand data centers can maintain their lowest interconnect latency to ensure maximum applications performance. Other non-Ethernet or proprietary network providers offer network-integrated or silicon-integrated gateways to Ethernet, which impose higher interconnect latency within the data center that can reduce the overall data center performance.

EXTEN Technologies and AIC to deliver NVMe-oF storage offerings optimized for supercomputing

AIC and EXTEN Technologies have partnered to provide a dense and scalable, software-defined NVMe-oF solution with record breaking performance, advanced management, provisioning, and RAID data protection services. This disaggregated storage based on the AIC Shark server provides industry leading price performance for HPC, analytics and ML/AI applications.

The AIC Shark server is ideal for dense computation, networking, and storage scalability. Four AMD EPYC processors manage storage processing for up to 24 local NVMe drives in its 2U enclosure. The AIC Shark server is an ideal scale-up (add JBOFs) and scale-out (add servers) platform for high-bandwidth shared storage applications that have a variety of data processing requirements.

The EXTEN HyperDynamic datapath running on Shark nodes creates a powerful data processing system with less than one microsecond of software overhead. This enables blistering performance with the lowest latency available. With simple striping, the 2U system achieves bandwidth of over 80GB/s. EXTEN supports both RDMA and TCP in parallel for maximum flexibility.

The EXTEN HyperDynamic software management suite offers a single REST interface to manage a cluster of NVMe-oF targets. This allows administrators to provision and manage the AIC Shark with the simplicity of a single machine, while maintaining the shared-nothing, linear scalability of NVMe-oF block storage performance. The EXTEN management framework is standards-based which ensures future-proof compatibility. Together, EXTEN software and the AIC Shark provide a perfect framework for implementing any reconfigurable, software-defined NVMe-oF solution.

Utilizing standard MPIO and Linux drivers, the Shark solution provides a robust storage platform built on standard tools that provides fault tolerance at both the target and drive level. EXTEN software ensures high performance resilience features including mirroring and dual parity RAID within each Shark node and volume replication across nodes without using client-side resources such as CPU, memory and I/O bandwidth.

Panasas releases PanFS on the ActiveStor Ultra Storage System to enhance performance, efficiency

Panasas announced that its re-engineered PanFS parallel file system, delivered on the Panasas ActiveStor Ultra appliance, is now shipping and in early deployment at Sumitomo Metal Mining and the University of Minnesota Supercomputing Institute. Built for price/performance and engineered for manageability, the new generation of PanFS is integrated with commercial off-the-shelf (COTS) hardware to deliver the fastest parallel file system at any price point.

The system is also deployed at a large U.S.-based aerospace manufacturer, at the world’s largest provider of subsurface data to the energy sector, and at a U.S.-based global security and aerospace company. 

The next generation of PanFS on ActiveStor Ultra offers unlimited performance scaling in 4 GB/s building blocks, utilizing multi-tier intelligent data placement to maximize storage performance by placing metadata on low-latency NVMe SSDs, small files on high IOPS SSDs and large files on high-bandwidth HDDs. The system’s balanced node architecture optimizes networking, CPU, memory and storage capacity to prevent hot spots and bottlenecks, ensuring consistently high performance regardless of workload.

The PanFS parallel file system is delivered on Panasas ActiveStor Ultra, a turnkey appliance that provides the extreme performance, enterprise-grade reliability and manageability required to process large and complex datasets. These datasets are associated with HPC workloads and emerging applications such as AI, precision medicine, autonomous driving, and augmented and virtual reality. 

ActiveStor Ultra goes from dock to data in one day, providing a plug-and-play solution that is easy to install, manage and grow. With ActiveStor Ultra’s modular architecture and building-block design, enterprises can start small and scale linearly by scaling metadata performance, bandwidth and capacity independently and without limitations.

“The re-engineered PanFS returns Panasas to the ranks of top-performing HPC data storage systems on the market, while maintaining the frustration-free storage experience our customers have come to expect and value,” said Faye Pairman, chief executive officer at Panasas. “With the new PanFS on ActiveStor Ultra, our customers can take advantage of the industry’s leading price/performance in a storage appliance that maximizes simplicity, boosts reliability, and delivers the lowest total cost of ownership.”

“The advanced-research computing infrastructure deployed at the Minnesota Supercomputing Institute has used Panasas data storage for more than 12 years,” said James Wilgenbusch, Ph.D., director of research computing, office of the vice president for research, University of Minnesota. “Ours is a truly mixed workload environment, and we value the ease and speed with which Panasas handles multi-discipline workloads such as CFD, CAE, computational physics and chemistry, and bioinformatics, to name a few. The new PanFS on ActiveStor Ultra expands on that tradition by delivering an appliance built for unlimited linear performance scalability at a highly attractive price point.”

“When we moved to a large cluster system we decided to deploy a parallel storage solution that delivered operational simplicity in addition to limitless scalability of capacity and performance,” said Toshiki Kishimoto, general manager, Computer-Aided Engineering and Development Dept., Technology Division, Sumitomo Metal Mining. “PanFS on ActiveStor Ultra offers us an environment that doesn’t require tuning to maintain peak performance, automates common maintenance and recovery tasks, and in case of questions, we have immediate access to support expertise from highly-qualified storage professionals.”

Kingston Technology releases NVMe SSD QoS benchmarks for HPC architectures

Kingston Digital Inc., a flash memory affiliate of Kingston Technology Company, demonstrated on Wednesday its server DRAM and latest NVMe PCIe Flash solutions for high performance computing (HPC) at Supercomputing 2019 (SC19). Supercomputing is an international conference that brings together the high-performance computing community. 

Kingston will demonstrate the consistent data performance and reliability of its forthcoming DC1000M U.2 NVMe SSD, purpose-built for data-intensive enterprise workloads and HPC architectures at the forefront of new artificial intelligence (AI) and machine learning (ML) applications.

During SC19, Kingston and Madison Cloud will demonstrate Private MultiCloud Storage (PMCS) architecture, providing a unified storage fabric for cloud and on-premise installations. 

Designed to keep data close to the Cloud without actually putting it in the Cloud, PMCS deploys the same storage platform wherever you are, providing the products and support that system builders need to move at the speed of Cloud. The two companies will showcase how users can run real-time MultiCloud workloads from a PMCS server in a MultiCloud architecture, built with Kingston DRAM, SATA and NVMe SSDs, providing consistent performance and a wide range of enterprise class features compared to cloud-native storage products.

Kingston will further demonstrate specific architectural advancements for Big Data and ML applications with the upcoming DC1000M U.2 NVMe SSD and recently launched DC500 SATA data center SSDs. 

The big data workflows validate possibilities for 10-bay NVMe server configurations while delivering 5.8 million IOPS at a sustained bandwidth of 23.8Gb/s. Kingston’s third demo presents ML application learning by naming thousands of satellite images per second as the DC1000M alleviates storage bottlenecks, accelerates large data sets and optimizes NVIDIA GPU performance for ML.

“Today’s supercomputers are powering a multitude of applications that are at the forefront of progress ― from oil and gas exploration to weather predictions, financial markets to developing new technologies,” said Keith Schimmenti, enterprise SSD business manager, Kingston. “These HPC applications require far greater consistency of performance and IO delivery. We are proud to share how Kingston’s solutions are an integral part of the stack behind the next breakthroughs in computing, providing the performance required in the world’s most extreme computing environments and applications.”