TY - GEN
T1 - ARMing x86 Games
T2 - 23rd ACM International Conference on Mobile Systems, Applications, and Services, MobiSys 2025
AU - Yen, James
AU - Wang, Jiarui
AU - Huang, Zhibai
AU - Wei, Zhixiang
AU - Zhang, Ziyang
AU - Chen, Chen
AU - Yu, Senhao
AU - Wang, Yun
AU - Wang, Hao
AU - Qi, Zhengwei
N1 - Publisher Copyright:
© 2025 Copyright is held by the owner/author(s). Publication rights licensed to ACM.
PY - 2025/9/25
Y1 - 2025/9/25
N2 - As ARM architecture becomes more prevalent in personal computers, users transitioning from x86-based Windows platforms face compatibility issues, particularly with x86 applications like games. Existing solutions, such as QEMU, Box64, and Apple's Rosetta 2, either incur high latency, face performance bottlenecks, or are limited to specific ecosystems. A key challenge remains the efficient translation of x86 status flags, which impacts performance.We propose a novel optimization method that enhances compatibility and performance by leveraging software-only strategies tailored to ARM hardware features. Using data flow analysis, our approach identifies when ARM's hardware flags can replace x86 flags, reducing reliance on software emulation and lowering translation overhead. This results in improved speed and compatibility for x86 applications on ARM, supporting demanding applications like games across x86 and ARM platforms without specialized hardware. Experimental results show significant performance gains, with computational tasks improving by up to 18%, and graphics rendering (FPS) also increasing by up to 18%. In particular, real-world testing on popular Steam titles demonstrates FPS improvements ranging from about 7% to over 12%.
AB - As ARM architecture becomes more prevalent in personal computers, users transitioning from x86-based Windows platforms face compatibility issues, particularly with x86 applications like games. Existing solutions, such as QEMU, Box64, and Apple's Rosetta 2, either incur high latency, face performance bottlenecks, or are limited to specific ecosystems. A key challenge remains the efficient translation of x86 status flags, which impacts performance.We propose a novel optimization method that enhances compatibility and performance by leveraging software-only strategies tailored to ARM hardware features. Using data flow analysis, our approach identifies when ARM's hardware flags can replace x86 flags, reducing reliance on software emulation and lowering translation overhead. This results in improved speed and compatibility for x86 applications on ARM, supporting demanding applications like games across x86 and ARM platforms without specialized hardware. Experimental results show significant performance gains, with computational tasks improving by up to 18%, and graphics rendering (FPS) also increasing by up to 18%. In particular, real-world testing on popular Steam titles demonstrates FPS improvements ranging from about 7% to over 12%.
KW - ARM
KW - binary translation
KW - EFLAGS
KW - x86
UR - https://www.scopus.com/pages/publications/105020385780
UR - https://www.scopus.com/pages/publications/105020385780#tab=citedBy
U2 - 10.1145/3711875.3729163
DO - 10.1145/3711875.3729163
M3 - Conference contribution
AN - SCOPUS:105020385780
T3 - MobiSys 2025 - Proceedings of the 23rd ACM international Conference on Mobile Systems, Applications, and Services
SP - 183
EP - 195
BT - MobiSys 2025 - Proceedings of the 23rd ACM international Conference on Mobile Systems, Applications, and Services
Y2 - 23 June 2025 through 27 June 2025
ER -