México se une a La Hora del Planeta con apagar luces de monumentos emblemáticos Spanish.xinhuanet.com

HPDA Is Now

According to IDC, 67% of the world's high-performance computing (HPC) data centers use AI and big data technologies. The accelerated convergence of
HPC, AI, and big data has led to the emergence of data-intensive high-performance data analytics (HPDA). But the explosion of data-intensive apps,
such as autonomous driving, genome sequencing, and weather forecasting, raises the requirements for real-time data analytics. And,
as computing power shifts from homogeneous to multi-cloud heterogeneous, workload models become more complicated, exposing
an urgent need for a solution that more efficiently and cost-effectively unlocks greater value from mass data.

Data Explosion

Complex Workload Models

Inefficient Analytics and Mobility

Global data is set to reach 180 ZB by 2025. More data needs to be stored for an extended period.

HPDA use cases like genome sequencing drive dozens of times of data growth.

Homogeneous computing power evolves into heterogeneous (GPU, ASIC, and FPGA), intensifying the challenge of diversified workloads.

Different service processing phases have different workload models, requiring one storage system to deliver high performance in both bandwidth and IOPS.

Data read/write performance does not match the computing power, making storage a bottleneck in analytical efficiency.

Accessing data across diversified protocols leads to inefficient data copying and time-consuming data migration.

The Trend of Data-intensive HPC, a white paper from Hyperion Research, highlights that in the transition from computing-intensive
to data-intensive workloads (typically big data, AI, machine learning, and deep learning), researchers, engineers, and data analysts
want to accelerate the time to results. For this, storage is crucial but it faces challenges and must be redefined.

Data Explosion

Global data is set to reach 180 ZB by 2025. More data needs to be stored for an extended period.

HPDA use cases like genome sequencing drive dozens of times of data growth.

Complex Workload Models

Homogeneous computing power evolves into heterogeneous (GPU, ASIC, and FPGA), intensifying the challenge of diversified workloads.

Different service processing phases have different workload models, requiring one storage system to deliver high performance in both bandwidth and IOPS.

Inefficient Analytics and Mobility

Data read/write performance does not match the computing power, making storage a bottleneck in analytical efficiency.

Accessing data across diversified protocols leads to inefficient data copying and time-consuming data migration.

The Trend of Data-intensive HPC, a white paper from Hyperion Research, highlights that in the transition from computing-intensive to data-intensive workloads (typically big data, AI, machine learning, and deep learning), researchers, engineers, and data analysts want to accelerate the time to results. For this, storage is crucial but it faces challenges and must be redefined.

Download White Paper

What Are

Data-Intensive HPDA Solutions?

HPDA involves data-intensive workloads that use HPC resources, like big data and AI workloads. It presents huge challenges, such as large data volumes, latency-sensitive processing, and complex algorithms. Data-intensive HPDA solutions provide a data infrastructure for supercomputing centers, research organizations, and enterprises, helping address the challenges from data-intensive apps’ development and unlock value and intelligence from their data.

Efficient Services

Data analytics with a super low latency meets the need for instant insights and shortens the R&D period. Data is seamlessly shared between development, testing, and production for efficient data mobility.

Simplified Management

A single storage system supports data mobility across protocols and provides unified resource O&M with high service agility, fulfilling the requirements for diversified computing power.

Optimized Costs

Intelligent tiering stores hot and cold data on different storage media to maximize data value and reduce footprint and energy consumption for optimal cost-effectiveness throughout the data lifecycle.

Benefits of

Data-Intensive HPDA Solutions

Data is now a strategic business resource, and data-intensive HPDA solutions are designed to support this. Centering on high-end storage, they provide an efficient, cost-effective platform for data-intensive HPC use cases like meteorology & oceanography, supercomputing centers, genome sequencing, and energy exploration.

Meteorology & Oceanography | More Accurate

As weather forecasting becomes ever more accurate, the data refresh frequency increases from once every 2 hours to once every few minutes, increasing the parallel computing capability requirements of data center storage. Our solution delivers:

Stable running of critical real-time services with high-end storage

Real-time, high-precision data analytics and processing through a hybrid-workload design

Supercomputing Center | More Intelligent

The shift from traditional to intelligent computing increases the performance from petascale to exascale, resulting in larger data volumes. To adapt to the complex multi-tenant workload models of supercomputing centers, storage must support cross-protocol access, resource isolation, and quality of service (QoS). Our solution provides:

Higher processing efficiency with dedicated HPC parallel file systems

Multi-protocol support that meets the access needs of various application systems and enables data mobility across HPC, big data, and AI platforms for lower total cost of ownership (TCO)

Higher management efficiency through full-lifecycle intelligent management

Genome Sequencing | More Efficient

Efficient genome sequencing helps cure diseases and develop new medicines. In the HPDA age, a single genome sequencer generates up to 6 TB of data every day, all of which must be stored long term. This requires high storage performance, stability, and cost-effectiveness. Our solution achieves:

Faster report output shortened from 13 years to just 1 day thanks to higher overall storage performance and quicker genome analysis and testing made possible by stable high bandwidth

Stronger genome analysis capabilities via quick hybrid analysis without data migration

Energy Exploration | More Cost-Effective

The evolution of seismic exploration from 2D to 3D and 4D results in 5- to 10-fold data growth. In seismic data processing, the aggregate bandwidth reaches 2 to 20 GB/s per PB and the latency per I/O must be 200 to 400 μs, requiring excellent storage scalability, bandwidth, and latency. More cost-effective energy exploration leads to a lower TCO, which is why our solution ensures:

Less physical storage space usage at a lower TCO with tiered storage

Worry-free, long-term data growth with one-time deployment and on-demand expansion

Meteorology & Oceanography

Supercomputing Center

Genome Sequencing

Energy Exploration

Meteorology & Oceanography | More Accurate

Stable running of critical real-time services with high-end storage

Real-time, high-precision data analytics and processing through a hybrid-workload design

Supercomputing Center | More Intelligent

Higher processing efficiency with dedicated HPC parallel file systems

Multi-protocol support that meets the access needs of various application systems and enables data mobility across HPC, big data, and AI platforms for lower total cost of ownership (TCO)

Higher management efficiency through full-lifecycle intelligent management

Genome Sequencing | More Efficient

Faster report output shortened from 13 years to just 1 day thanks to higher overall storage performance and quicker genome analysis and testing made possible by stable high bandwidth

Stronger genome analysis capabilities via quick hybrid analysis without data migration

Energy Exploration | More Cost-Effective

Less physical storage space usage at a lower TCO with tiered storage

Worry-free, long-term data growth with one-time deployment and on-demand expansion

Why Huawei?

Huawei has 12 R&D centers with over 4000 R&D personnel and 3000 patents, all dedicated to storage, serving more than
15,000 customers in over 150 countries. With ultra-high density and multi-protocol interworking capabilities, Huawei HPDA solution is
designed for hybrid workloads to unlock data potential and tackle the challenges arising from data-intensive apps.

Hybrid workloads-oriented

A single storage device delivers industry-leading 32 GB/s bandwidth and 400,000 IOPS per U, boosting service efficiency.

Multi-protocol interworking

Seamless interworking between POSIX, NFS, CIFS, HDFS, S3, and MPI-IO with zero data copying is purpose-built for HPDA scenarios, simplifying data management.

Ultra-high-density design

24 disks/U, high-density, large-capacity hardware and 16 disks/U, high-density, NVMe all-flash hardware combined with intelligent data tiering offer optimal cost-effectiveness.

Learn More

Newly Released OceanStor Pacific Series

See Who's Already Taken the First Step to Success

Using Mass Data to Open a New Era of Energy Exploration for Sinopec Geophysical Research Institute

Oil and gas exploration involves complex data processing. How to solve the problems found in this procedure is critical to oil and gas exploration.

Constructing the World's First SKA Prototype with Huawei OceanStor Mass Data Storage

Shanghai Astronomical Observatory released the first SKA regional center prototype at the SKA Engineering Meeting in November 2019.

A deep-sea mining platform in the middle of the ocean on a clear day, to represent Huawei oil and gas exploration solutions

Exploration Company Improves Storage Architecture for Success

According to IDC research, one-fourth to one-third of the value generated by a typical oil and gas exploration and production company’s annual activities is owed to data. Lundin Norway is no exception.

Find the Right Product and Solution for You

OceanStor Pacific 9950

High-density performance model
5 U, 8 nodes
80 NVMe SSDs
160 GB/s bandwidth, 2 million IOPS

Learn More

OceanStor Pacific 9920

Performance model
1 node per 2 U chassis
All-flash distributed storage
Superior performance and flexible component configurations

Learn More

OceanStor Pacific 9550

High-density capacity model
5 U, 120 x 3.5-inch SATA disks
1.68 PB of raw capacity
Elastic EC

Learn More

OceanStor Pacific 9520

Capacity model
1 node per 2 U chassis
Distributed storage
Multiple device types and flexible component configurations

Learn More

OceanStor Pacific 9540

Capacity model
1 node per 4 U chassis
Distributed storage
High capacity density and flexible component configurations

Learn More

Resources

OceanStor Pacific Series Data Sheet

Presentation of OceanStor Pacific Series for HPDA

OceanStor Pacific 9550 Hardware Introduction

谷维素治什么病	小儿疳积是什么症状	孕妇为什么要躲着白事	右耳朵发烫是什么征兆	jk制服是什么意思
什么人容易得胆汁淤积	xxoo是什么	什么地流着	被蛇咬了挂什么科	泞字五行属什么
女性私处痒是什么原因引起的	mario是什么意思	什么的玉米	鱼爱吃什么食物	心脏早博是什么意思
面首是什么	藏医最擅长治什么病	血小板偏高是什么意思	人大代表是什么	赘是什么意思

为什么会脑供血不足yanzhenzixun.com	专科考研需要什么条件hcv9jop2ns7r.cn	3岁小孩说话结巴是什么原因wuhaiwuya.com	支气管哮喘是什么原因引起的hcv9jop8ns0r.cn	尿常规是检查什么的dayuxmw.com
阳绿翡翠属于什么级别hcv8jop8ns1r.cn	mt是什么单位hcv7jop6ns9r.cn	第一次同房是什么感觉baiqunet.com	感染是什么症状hcv7jop4ns6r.cn	肠胃消化不好吃什么药hcv8jop8ns0r.cn
容易长口腔溃疡是什么原因hcv8jop5ns8r.cn	梅毒通过什么传播0297y7.com	五花肉和什么菜炒好吃hcv7jop9ns8r.cn	奶奶的姐姐叫什么hcv7jop6ns4r.cn	老年人缺钾是什么原因引起的gangsutong.com
茹字五行属什么hcv9jop7ns9r.cn	泽五行属什么hcv8jop2ns2r.cn	吃什么可以补钾luyiluode.com	瘦人吃什么长胖hcv9jop6ns1r.cn	夏天喝什么汤好hcv9jop6ns9r.cn

Enterprise

Huawei Cloud

Opérateurs

Grand public

Corporate