High-performance computing, graph analytics key to IT's transformation

More efforts are needed to explore the application of high-performance computing and graph analytics, as the combination of the two cutting-edge technologies is likely to become a key driving force influencing future transformation of the entire IT computing sector, experts said.

High-performance computing has been used in carrying out scientific research and development, such as weather forecasts. It now can be used in more areas such as graph analytics, which can help deliver enhanced value. High-performance graph computing is the core technology of artificial intelligence and big data, said Zheng Weimin, an academician at the Chinese Academy of Engineering.

Graph analytics is a set of analytic techniques that help explore relationships among organizations, people and transactions.

Market research company Gartner said in a report that by 2023, graph technologies will facilitate rapid contextualization for decision-making in 30 percent of organizations worldwide, and its average annual growth rate is forecast to exceed 100 percent.

Graph analytics can help data and analytics leaders find unknown relationships in data and review data not easily analyzed with traditional analytics. For example, as the world scrambles to respond to the ongoing COVID-19 pandemic, graph technologies can relate entities across everything from geospatial data on people's phones to facial-recognition systems that can analyze photos to determine who might have come into contact with individuals who later tested positive for the coronavirus, Gartner added.

"More efforts are needed to develop high-performance graph computing technology. The software and hardware involved in this technology are areas where China is likely to face restrictions from others in the future," Zheng said.

Yang Juan, president of Haizhi Network Technology Beijing Co Ltd, a Beijing-based graph technology company, said graph computing has already been used in sectors such as finance and energy in China.

Experts said graph analytics, for instance, is being used to help identify fraud and money laundering activities and give warnings to financial sector players. When it comes to e-commerce, graph analytics can also help offer more accurate online real-time product recommendations. In cybersecurity, the technology can achieve real-time identification of cyberattack perpetrators, and protect information security.

Wu Yongwei, a professor at the department of computer science and technology at Tsinghua University, said although high-performance computing and graph computing technologies have been developed in China for some time, the existing solutions still cannot meet the needs of complex types of operations, and there is not good synchronicity between hardware and software. A broader application market needs to be urgently developed.

To help quicken the development of high-performance graph computing in China, Haizhi has partnered with Tsinghua University to jointly build a workstation in the sector.

"We aim to build a complete set of high-performance graph computing ecosystems and industrial clusters including hardware, software and industry application services to help build China's core industrial base for high-performance graph computing," Yang said.

She added that Haizhi has been dedicated to serving governments, financial institutions and energy companies with graph technologies for some time. The company's technologies had been used to prevent the spread of the COVID-19 pandemic in more than 10 provinces in China in 2020. The company is now exploring how to widen the use of its digital technologies and solutions in smart transportation and smart city initiatives.