BEIJING, April 2, 2024 — The 2024 ASC Student Supercomputer Challenge (ASC24) finals are scheduled to take place at Shanghai University from April 9 to 13, 2024. Among the over 300 teams representing universities worldwide, only 25 have progressed to the finals.

Throughout the competition finals, teams will face a series of challenges including optimizing large language model (LLM) inference, tackling tasks like OpenCAEPoro and GoMars, as well as competing in high-performance benchmarks such as HPL and HPCG.

Additionally, teams will engage in group competitions and tackle a mystery application. Each team will present their projects before the jury, contending for prestigious awards including the Champion, Silver Prize, e Prize, and the Highest LINPACK.

The ASC24 Student Supercomputer Challenge Finals

ASC is committed to assisting students in addressing state-of-the-art scientific challenges through the utilization of the most advanced computing products and technologies. This endeavor aims to explore emerging research paradigms and confront the various challenges humanity faces. The focal point of the ASC24 finals will revolve around tackling cutting-edge scientific issues, such as advancing large language models (LLM), exploring Mars, and enhancing techniques for oil and gas exploitation.

The prestigious e Prize will be bestowed upon the winners of the OpenCAEPoro task, which entails simulating the seepage of multiphase fluids, such as oil, gas, and water, particularly within contexts relevant to petroleum extraction. Competitors are tasked with skillfully optimizing large-scale parallel computing processes, emphasizing enhancements in computing performance and parallel efficiency of discrete algorithms.

In the GoMars task, participating teams are challenged with optimizing parallel communication and refining computational load balancing within the dynamic framework of GoMars. GoMars represents a new generation of Mars atmosphere models. The objective is to enhance operational efficiency and achieve more precise predictions of the Martian atmosphere.

In the LLM Inference Optimization Challenge, participating teams are assigned the task of constructing and refining inference engines utilizing the widely employed open-source LLM, AquilaChat2-34B. The primary objective is to expedite inference processes by employing 4-bit or less quantization while maintaining overall accuracy.

Alongside the primary challenges, the finals will include a Group Competition aimed at fostering cross-team collaboration among students. Students will be randomly allocated to one of five groups. Each group will work collectively on the ParaSeis task, employing ParaSeis to simulate seismic wave propagation and refine the algorithm through collaborative communication. The group’s performance will be evaluated based on the average score of its constituent teams, with the group achieving the highest score declared as the victor.

The finals will bring together formidable teams from renowned universities, including Peking University – the champion of the 10th ASC onsite finals, the Chinese University of Hong Kong – the champion of the 10th ASC virtual finals, University of Science and Technology of China, Huazhong University of Science and Technology, Shanghai Jiao Tong University, and Friedrich-Alexander-University Erlangen-Nuremberg.

Numerous new teams are set to join the finals, with Shanghai University, Lanzhou University, and Southern University of Science and Technology among them. Excitement surrounds the debut of teams from the National University of Cordoba, Macau University of Science and Technology, and Southwest Petroleum University, making their inaugural ASC finals appearance this year. The enthusiasm of these new contenders promises to elevate the competition, ensuring an intense battle until the very end.

About ASC

The ASC Student Supercomputer Challenge is the world’s largest student supercomputer competition, sponsored and organized by the Asia Supercomputer Community with support from experts and institutions across Asia, Europe, and America. The main objectives of ASC are to encourage the exchange and training of young supercomputing talent from different countries, improve supercomputing applications and R&D capacity, boost the development of supercomputing, and promote technical and industrial innovation. The first ASC Student Supercomputer Challenge was held in 2012 and has since attracted over 10,000 undergraduates from all over the world. Learn more about ASC at http://www.asc-events.net/StudentChallenge/index.html.

---

Source: ASC