Huawei proposes a new storage concept

At the 2022 Innovative Data Infrastructure Forum, Huawei proposed a new, innovative storage concept of "building a data-centric, trustworthy storage foundation for diverse applications". This strategy aims to continuously build highly reliable and performant fundamental storage technologies. By prioritizing the decoupled storage-compute architecture and diverse data application acceleration engines, Huawei is set to help enterprises accelerate their digital transformation and cope with diverse data applications.

  • 1 year ago Posted in

In our age of huge data growth, storage technologies face the following changes:

1. Innovation for future workloads of emerging applications, such as distributed database, big data, AI, and High-Performance Data Analytics (HPDA) applications.

2. Acceleration for data services, which requires faster and real-time data analysis and processing to improve production efficiency and user experience.

3. Data protection at a higher standard to mitigate against frequent data attacks that could cause huge economic loss.

4. Green solutions where the data storage must be equipped with innovation and high energy efficiency.

Dr. Peter Zhou, the Vice President of Huawei and President of Huawei IT Product Line, explained that Huawei has already offered a full series of enhanced competitive storage products that proactively addressed these changes, including OceanStor Dorado All-Flash Storage, OceanStor Pacific Distributed Storage, OceanProtect Backup Storage, and FusionCube. These offerings are well suited to key industries and enterprises. But that is not enough to build a data-centric, trustworthy data storage foundation for diverse applications, if there are no further efforts in the following three aspects.

First, the decoupled storage-compute architecture is a must. As enterprises continuously accelerate their digital transformation, emerging data applications are gradually becoming production applications, demanding higher data reliability. Furthermore, the lifecycle gap between computing and data is expanding, requiring flexible and independent management and maintenance of computing and storage resources. The decoupled storage-compute architecture will run on elastic, reliable, and cost-effective storage to facilitate faster, more reliable, and efficient services at lower costs.

Second, there should be more data application acceleration engines within storage systems. To cope with diverse data applications, future-oriented storage is expected to not only house data persistently, but also be used to build data infrastructure that combines a data persistence layer with the application acceleration engine. Emerging data applications such as distributed databases, big data, and AI applications that produce mass data continue to grow. Future storage systems will be built with metadata management and intensive data processing to form diverse data application acceleration engines, improving end-to-end processing efficiency by up to 10 times and significantly improving customer experience and operational efficiency.

And finally, our green design, essential to the product development, covers three aspects: green production, such as renewable materials; green products with low-carbon components, like NAND flash, high-density hardware design, leading data reduction, and resource pools for less space occupied and better resource efficiency; and green enablers, that is, supporting enterprise services to optimize production operations in an energy-efficient way.

Dr. Peter Zhou also noted over three decades, data storage has evolved to become the optimal foundation of high-value data in line with the data application development. "We are ushering in the yottabyte era," He noted, adding "Data applications are growing faster than ever. This forum's theme is Green, Acceleration, and Innovation, and this is also the development orientation of Huawei Storage. Together with our partners, we will provide high-quality storage products and solutions to create more value for customers."

New tape innovation from Quantum delivers infrastructure for AI with maximum storage density,...
Powering a new era of computing, NVIDIA has unveiled the NVIDIA Blackwell platform — enabling...
Eviden has won a new contract with the Danish Centre for AI Innovation, owned by the Novo Nordisk...
Next phase of NVIDIA Hybrid AI collaboration places transformational AI in the hands of every...
Vertiv brings its high-density power and cooling infrastructure expertise to the NVIDIA Partner...
Initiative involves development of AI frameworks to expedite time to market and reduce the...
DDN has demonstrated its reference architecture, which integrates NVIDIA BlueField-3 DPUs into DDN...
Supermicro is announcing its latest portfolio to accelerate the deployment of generative AI. The...