10-Member Team Drawn from Chinese Institutions Wins 2026 Gordon Bell Prize

A Chinese ten‑member team won the 2026 ACM Gordon Bell Prize for delivering the first exascale classical simulation of the nitrogen‑fixation FeMo‑cofactor.

10-Member Team Drawn from Chinese Institutions Wins 2026 Gordon Bell Prize

Chinese Team Wins 2026 Gordon Bell Prize for Exascale Nitrogen‑Fixation Simulation

A collaborative team of ten researchers from several Chinese institutions has been awarded the 2026 ACM Gordon Bell Prize. The prize recognises their breakthrough in using exascale supercomputers to solve a long‑standing challenge in biochemistry: a classical, high‑performance simulation of the iron‑molybdenum cofactor (FeMo‑cofactor) that underpins nitrogen fixation.

Why the FeMo‑cofactor matters

Nitrogen fixation—converting atmospheric N₂ into biologically usable ammonia—is essential for plant growth and, by extension, global food production. The FeMo‑cofactor (FeMoco) is the active site of the nitrogenase enzyme that drives this reaction. Understanding its electronic structure and dynamics requires quantum‑chemical calculations that are notoriously computationally intensive. Historically, researchers have relied on approximations or smaller‑scale simulations, limiting insight into the detailed mechanisms.

Achieving exascale performance

The winning team developed a bespoke codebase that was tuned for two Chinese exascale machines, LineShine and the New‑generation Intelligent Super‑computer (CNIS). On 20,480 nodes of LineShine, the application sustained 1.456 EFLOP/s in double precision (about 59 % of the machine’s peak) and 5.626 EFLOP/s in mixed precision. These figures represent the first reported instance of a classical (non‑quantum) simulation of the FeMo‑cofactor reaching sustained exaflop performance.

Key technical innovations included:

  • Domain‑specific optimisations for the electronic‑structure algorithms used in the simulation, reducing communication overhead across nodes.
  • Mixed‑precision strategies that leveraged lower‑precision arithmetic where tolerable, boosting throughput while preserving scientific accuracy.
  • Scalable I/O pipelines to handle the massive data volumes generated by the simulation, ensuring that storage bandwidth did not become a bottleneck.

Implications for science and industry

By delivering an exascale‑level classical solution, the team demonstrates that today’s most powerful supercomputers can tackle problems previously thought to require quantum hardware. The detailed electronic‑structure data generated could inform the design of synthetic catalysts that mimic nitrogenase, potentially leading to more energy‑efficient fertilizer production.

Moreover, the achievement showcases the growing capability of Chinese high‑performance computing (HPC) infrastructure and its impact on global scientific research. The prize, presented at the SC26 conference in Chicago, highlights the importance of interdisciplinary collaboration—combining expertise in chemistry, computer science, and HPC engineering—to push the boundaries of what classical computation can achieve.

Looking ahead

While the current work focuses on a specific biochemical system, the methodologies developed are broadly applicable to other quantum‑chemical challenges, such as materials discovery and drug design. Future work will likely explore further optimisation, scaling to newer exascale platforms, and integration with emerging quantum‑computing approaches to address even larger, more complex systems.

--- The article is based on the ACM press release and the team’s abstract describing their exascale performance.

Hype check

Partly

The claim of a first exascale classical solution is supported by performance numbers in a press release, but lacks peer‑reviewed validation or independent verification.

  • Peer-reviewed paper, or press release only?Press release only — Only a press release and abstract are cited, no peer‑reviewed paper.
  • Error-corrected (logical) qubits, or physical qubits only?Logical / not a qubit claim — The story does not involve qubit counts.
  • A new milestone, or a rebrand / restatement?New — First reported sustained exaflop classical simulation of FeMo‑cofactor.
  • Independent confirmation?Not yet — No external confirmation or replication mentioned.
  • Does the headline claim match what the source shows?Matches — Headline accurately reflects the prize for the exascale achievement.

An editorial assessment of the claim as reported, not a judgment of the company or researchers. How we rate hype →

Source

Read original source

Why we picked this

Accepted from curated QuantumUnzipped source (relevance scoring disabled).

← Back to all articles