Welcome to the arcSYSu lab

Architecture and Systems Upscaling @ Sun Yat-sen University!
HPC/AI Research led by Xianwei (ARX)!

arcSYSu, abbreviated as ARX (pronounced /ɑːrks/), is a research group working under the guidance of Associate Professor Xianwei Zhang in the School of Computer Science & Engineering at Sun Yat-sen University. Our research spans computer architecture, compilation and runtime, scalable systems, and computing platforms, with the mission of building next-generation HPC/AI infrastructure. At arcSYSu, we design and build scalable, sustainable and systematic (S3) foundations that connect architecture, system software, and applications, transforming research innovations into practical computing infrastructure.

arcSYSu is proudly part of the Interdisciplinary Research Center (xRC), jointly directed by Prof. Yutong Lu and Prof. Nong Xiao. We work closely with other exceptional teams across the center on a broad spectrum of HPC+AI design and application areas.

Our research is supported by prestigious programs, including the National Key R&D Program, the National Natural Science Foundation of China, as well as collaborative partnerships with industry leaders like Tencent®, Huawei®, Xiaomi® and Phytium®.

We are always seeking passionate PhD/MS students, postdoctoral researchers and undergraduates to join us in exploring the frontiers of computing! (For more details, please refer to our ↗ FAQs).

News

07/2026: [Paper] Two papers (HSPref and coMulator) accepted to MICRO'2026. Congrats to Han and Yuhao!

07/2026: [Talk] One invited talk on AgentSys delivered at Huawei's AI Compilation Workshop!

07/2026: [People] Five ARXers graduated! Cheers and all the best!

06/2026: [Paper] New preprint on Agent Assessment (RAMP) released on arXiv.

05/2026: [Paper] New preprint on Agent Skill Compilation (SkCC) released on arXiv.

05/2026: [Fund] One grant awarded for LLM acceleration! Thanks!

05/2026: [Paper] One paper (SMEAtten) accepted to Euro-Par'2026! Congrats to Tengyang!

04/2026: [Fund] Two grants awarded for GPU computing, and agentic systems! Thanks!

03/2026: [Tool] New YatCC platform launched for AI-powered Teach+Research!

12/2025: [Teach] Two Provincal Teaching Awards received! Thanks all!

12/2025: [Paper] New preprint on Compilation-assisted Emulation (coMulator) released on arXiv.

11/2025: [Paper] One paper (Bullet) accepted to ASPLOS'2026! Congrats to Zejia!

... see all News