The constant advancement in technology has led to a lot of breakthroughs in the industry. Over the last few years, high performance computing (HPC) has been an essential element in the progress of scientific research, enabling significant changes in several disciplines such as climate modeling and national security.
The rapid advancement and widespread adoption of artificial intelligence (AI) has expanded the limits of research and discovery, resulting in the establishment of systems and functions that have become part of our daily lives.
It is predicted that global high performance computing (HPC) market revenues are likely to experience considerable growth in 2023, amounting to more than USD 35 billion. By 2026, HPC market revenues worldwide are forecast to rise to USD 40 billion.
In Asia, the adoption and development of HPC has been constantly growing due to the region's fast economic growth, significant investments in research and development (R&D), and the growing demand for enhanced computational capabilities.
Understanding High Performance Computing
HPC is a combination of computing resources designed to achieve performance levels that surpass what can be done by a single workstation, server, or computer. This may appear as either specialized supercomputers or as collectives of individual computers operating in harmony. HPC systems can be implemented either on-site, via a cloud-based infrastructure, or through a hybrid model that integrates both environments.
Every individual machine within an HPC cluster, known as a node, carries out distinct tasks as part of the total calculation. These nodes are classified into controller nodes, responsible for managing crucial services and coordinating tasks amongst nodes; interactive or login nodes, where users access the system; and compute or worker nodes, which carry out the computations. HPC systems can efficiently handle activities that are either too massive or time-consuming for ordinary computers by executing algorithms and software updates simultaneously across multiple nodes.
The essential components of HPC systems include computing, storage, and networking. Computing includes processing units responsible for doing calculations, storage involves the data servers that capture the output, and networking facilitates the efficient exchange of results between these processing units. To ensure efficiency in HPC systems, it is essential to have high speed disks, fast memory, and low-latency, high-bandwidth networking. These components are crucial for managing the complexity and scale of HPC projects effectively.
Also Read: Empowering Smart Cities in Asia with Edge Computing
Adopting HPC in Asia
The bustling economy of Asia creates a favorable environment for the implementation of HPC. Countries in Asia are among the frontrunners in utilizing HPC systems to enhance their scientific research, industrial production, and technological innovation.
In 2022, OneAsia introduced Hong Kong’s pioneer HPC solution, OAsis. It was expected to transform data processing and simulation in the country, thereby facilitating progress in academic research and industry innovation. It was also established to boost Hong Kong's position as a global hub for cutting-edge technology but also corresponds with wider national and regional technical objectives.
In the same year, the National Development and Reform Commission of China (NDRC) announced its plan to build eight national computing hubs and ten national data-center clusters in the country, including the establishment of a hub in the Greater Bay Area (GBA).
Meanwhile, South Korea has taken the initiative to bridge the technological gap with global leaders by pursuing the development of an exceptionally powerful supercomputer specifically designed for industrial applications.
Telecom Review Asia Feature: Exploring Infinite Possibilities with Quantum Computing
South Korea, ranked as the fourth-largest economy in Asia, has historically been a center of technological advancement. In 2023, the Ministry of Science and ICT unveiled a national strategy to boost HPC capabilities, which includes improving supercomputers and exploring quantum computing, recognizing the significance of HPC as an essential infrastructure in today’s digital landscape.
Reports indicate that South Korea plans to develop its seventh supercomputer in 2025, which will have an exaflop of performance, meaning it will be able to perform at least one quintillion floating-point operations per second.
High-performance computing (HPC) is on the rise in Asia, driven by increasing cyberthreat levels and the need for enhanced technological capabilities. Vietnam leads Southeast Asia in cyberthreats, prompting initiatives like the National Supercomputing Centre (NSCC) to adopt advanced technologies to secure its HPC systems and networks.
In Malaysia, Petronas Digital partnered with Microsoft and Cegal to build the Microsoft Azure HPC AI infrastructure platform, optimizing operational efficiency in Petronas Upstream’s applications.
Moreover, according to Tan Tin Wee, Chief Executive, National Supercomputing Centre (NSCC), Singapore’s A*STAR’s Institute of High-Performance Computing (IHPC) supports over 300 researchers who actively engage in scientific simulations, gather data analytics, and conduct AI research.
Singapore also hosts the annual Supercomputing Asia conference (SCA’23), collaborating with Australia, Japan, and Thailand, and has inaugurated the Alliance of Supercomputing Centres (ASC) to foster resource sharing.
Additionally, Huawei's CloudFabric 3.0 Hyper-Converged DCN solution maximizes computing power with lossless ethernet, further advancing HPC capabilities in the region.
Read More: UNICOM Engineering Boosts Computing Power in Collaboration with Intel and Dell
Closing Note
Despite its advantages in advancing industries in the region, substantial obstacles remain, which need to be resolved in order to fully realize HPC’s potential. A significant obstacle is the prevalence of high expenses related to infrastructure, including hardware, software, and maintenance, which can be seen in smaller academic institutions and companies.
To counteract this, the government and business sectors could collaborate to explore cost-efficient solutions, such as cloud-based HPC services, to improve accessibility. In addition, the lack of a skilled workforce with expertise in parallel programming and data processing is hindering the use of HPC. Furthermore, data privacy and security concerns in industries such as healthcare also require strong cybersecurity measures and adherence to regulations to safeguard sensitive information.
HPC can be an effective tool that can accelerate scientific, technical, and business advancement. Its ability to facilitate the efficient analysis of large datasets allows organizations to address intricate challenges and shape a more defined future. With the ongoing advancement of technology, the significance of HPC will likely increase, marking its position as a fundamental element of modern technology.
Telecom Review Asia Analysis: AI, HPC, and Quantum Technologies Poised to Boost Innovation
Read More: SoftBank Corp. Launches Top-level Generative AI Computing Platform
In-Depth Market Insights: Cloud and Edge Computing Transforming Industries in Asia Pacific
