New ML approach accelerates semiconductor device simulation by 10,000x

Researchers have developed a novel approach for simulating semiconductor devices by combining machine-learned electronic structure models with a quantum transport solver. This new framework offers a significant speedup, up to 10,000 times faster than traditional density functional theory, enabling the simulation of devices with over 20,000 atoms while maintaining high accuracy. The method was applied to study molybdenum disulfide (MoS2) based field-effect transistors, revealing that undercoordinated metal atoms at the semiconductor-oxide interface critically influence the electronic current. AI

IMPACT This accelerated simulation technique could speed up the discovery and design of new semiconductor devices.

RANK_REASON The cluster contains an academic paper detailing a new computational method for materials science research. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.LG →

• aluminium oxide
• density functional theory
• Field-effect-transistors
• hafnium(IV) oxide
• Hamiltonian matrix
• Manasa Kaniselvan
• molybdenum disulfide
• quantum transport

• paper
• infra

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →