Traditional Culture Encyclopedia - Traditional festivals - Climbing a mountain higher than Mount Everest is a breakthrough in data storage technology.

Climbing a mountain higher than Mount Everest is a breakthrough in data storage technology.

Text: Yu Hongtao Science and Technology Business.

In the material world, the mountain peak is a symbol of natural forces; In the digital world, data is the source of wisdom.

Perhaps because of this, Huawei named its scientific research award focusing on data basic technology Olympus Mons, which is specially used to reward researchers who have made technological breakthroughs in data storage.

Setting up awards is just a formality. Behind the Olympus Prize is Huawei's ambition to promote the breakthrough development of data technology by pooling the capabilities of all parties in Industry-University-Research, thus providing better data infrastructure for the development of digital economy.

With the advent of the digital age, the value of data has become more and more prominent, and it has become the core asset of the country, enterprises and even individuals.

Compared with the traditional economy, the essence of digital economy is the circulation of data, and data has become the main production factor of intelligent society. According to IDC's research, in 2020, global * * * created 59ZB of data, which will reach 163ZB by 2025.

Such a huge data asset needs to go through data collection, data storage, data analysis and other processes to produce value, of which data storage is undoubtedly the foundation. In the data center, storage, computing and network have become the key infrastructure to support the whole digitization process.

With the rapid growth of data, the forms of data are increasingly diversified, and unstructured data such as video, pictures and audio become the main body of data. If these complex data want to give full play to its value, it needs more efficient data storage and data management.

According to statistics, only 2% of the data has been preserved, and only 10% of the preserved data has been analyzed and utilized. Zhou Yuefeng, president of Huawei's data storage and machine vision product line, said that data is playing an increasingly important role in the digital transformation of enterprises. However, enterprises are faced with the problem that massive data cannot be saved, flowed and managed well.

In order to meet the growing data storage needs of customers, Huawei advocates building end-to-end data capabilities, including computing, storage, utilization and AI capabilities, so as to maximize the value per bit and optimize the cost per bit in the whole life cycle of data.

Huawei's efforts have achieved results. Today, more and more governments and enterprises use Huawei's data storage solutions to manage data assets.

Gansu Dunhuang Research Institute is using Huawei's massive storage solution to realize the permanent preservation and sustainable utilization of Dunhuang grottoes cultural relics through computer technology and digital image technology.

However, there are 735 caves, 45,000 square meters of murals and 24 15 clay sculptures in Dunhuang Mogao grottoes. Digitizing so many cultural relics and realizing the goal of building a digital Dunhuang museum means a lot of investment and massive storage equipment. Obviously, to solve this problem, it is not enough to rely solely on Huawei's own efforts, but it needs the participation of all parties to achieve a new breakthrough in storage technology by building an industrial technology ecology. This is also the original intention of Huawei to set up the "Olympus Award".

It is understood that Huawei's "Olympus Award" focuses on two major technical problems in the data field every year to find solutions. At the Global Data Storage Professors' Forum at the end of last year, the second session of "20021Year Olympus Hanging Red" was decided: one is to create the most cost-effective data storage per bit, and the other is to achieve the root technology breakthrough of the next generation storage industry. For each question, Huawei has given up to 6.5438+0 million yuan.

Huawei hopes that through the establishment of the "Olympus Prize", it will work together with the academic circles in the technical directions of cloud-oriented cloud storage services, new data-centered application storage systems, AI-driven storage software architecture, and innovative system architecture to build a better data storage system.

As we all know, the main factor that hinders the popularization of electric vehicles is the energy density of batteries, which determines the availability of electric vehicles. In the data center, the storage density of data will become the core challenge in the future, which determines the quality of our intelligent society.

Scientists have defined the next development goal: to realize data storage with 100x performance density and 100x capacity density with limited resources. In order to improve the storage capacity, compression algorithm is one of the core technologies, which can reduce the storage cost of data and help users alleviate the cost pressure caused by the explosive growth of data scale.

However, as a difficult problem in storage technology, compression algorithm has not achieved outstanding results for many years.

In order to break through the bottleneck of compression algorithm and stimulate the vitality in the field of data compression, since 2020, Huawei has cooperated with Moscow State University to hold a global data compression competition to promote the research of data compression root technology.

This year's second global data compression competition invited prestigious technical experts in the field of compression as judges; The use of high-performance computing data such as electron microscope and remote sensing is closer to the frontier and actual scene. The contest designed five kinds of data sets (events): quantitative data compression, qualitative data compression, mixed data compression, small data compression and entropy coding optimization.

At the same time, the competition also added a relatively low-difficulty coding algorithm optimization project for college students to attract more campus algorithm experts to participate in the competition. In the award setting, multi-dimensional incentives are further reflected, and leadership awards, special awards and student participation awards are added.

This year's data compression contest officially started on June 15. Entries will be accepted by the end of June 1 1, and the winning results will be announced at the end of February 1. By mid-July, the organizing committee of the competition had received nearly 80 applications from all over the world only 1 month after the start of the competition.

With the Olympus Award and the Global Data Compression Competition entering the second session, "Olympus" has become a new brand that Huawei is striving to build, which is specially used to strengthen cooperation in Industry-University-Research and jointly promote the progress of the data storage industry with the academic community.

From the winners of the first Olympus Prize, we saw the breakthrough of scientific research in the field of data technology innovation.

The "Construction and Key Technologies of Persistent Memory Storage System" by Tsinghua University Shu, a team that won the million-dollar prize, innovatively proposed the design method of persistent memory file system and key value storage and the distributed persistent memory sharing framework, which overcame a series of problems such as data structure, memory management, consistency and security, and solved the problem of efficient data storage based on new storage media.

In addition, Chen Rong of Shanghai Jiaotong University's "Efficient Data Processing System Based on New Heterogeneous Hardware" and Feng Dan of Huazhong University of Science and Technology's "Efficient and Reliable NVM Technology" are also innovative and advanced, with industrial value and application prospects.

Similarly, in the first global data compression competition, many remarkable achievements emerged.

For example, the pglz algorithm designed by the award-winning player Peter Thamm breaks the general understanding of the compression ratio and performance of the fast compression algorithm and guides the optimization direction of the compression algorithm; Konstantinos Agiannis's entry algorithm, in the text scene test, both compression rate and compression performance exceed the benchmark algorithm recognized by the industry; Andreas Debski's fast image compression algorithm has reached the compression ratio of 120%, which is recognized as the benchmark algorithm in the industry, showing his profound knowledge of image compression algorithm.

The success of the past year has also made us have higher expectations for this year's "Olympus". Climbing the highest mountain in the solar system means that the participants in the whole field of data storage technology can unite as one for the first time, * * * take basic theoretical research as the direction, break through key technical problems, accelerate the industrialization of scientific research achievements, and realize win-win cooperation in Industry-University-Research.

In this process, on the one hand, Huawei has played the role of an industry leader and vigorously promoted the cooperation process in Industry-University-Research; On the other hand, we also actively participated in it, and made system innovation around the media, network, architecture and management of the next generation storage through five innovative laboratories, such as data fiber, intelligent storage, memory storage, data restoration and video storage, and through the cooperation of more than 4,000 R&D engineers Qi Xin.

We also have reason to believe that through global and all-field collaborative innovation, we will surely usher in a breakthrough in data storage technology, achieve better data storage efficiency through technical reconstruction, and let the world enjoy the dividend of digital technology, thus promoting the intelligent upgrade of thousands of industries.