Two Papers Accepted by MICRO 2026 and One Paper Accepted by ICCAD 2026
Two papers on privacy-preserving AI accelerators (i.e., OptiPrime and Helios) are accepted by MICRO 2026. One paper on efficient AI (i.e., MatMoE) is accepted by ICCAD 2026.
The accepted papers are listed below:
- OptiPrime: Optimizing Private Inference through Protocol-Hardware Codesign (led by Jiangrui Yu, collaborated w/ Si Chen and Junfeng Fan from OpenSecurity Research)
- Helios: Melting Kernel Boundaries for GPU-Accelerated HE via Graph Rewriting and Microarchitecture-Aware Mapping (led by Yi Chen and Ziyu Tang, collaborated w/ Mingzhe Zhang from Ant Group)
- MatMoE: Matryoshka Mixture-of-Experts with Dynamic Mixed-Precision Quantization for Efficient Inference (led by Renjie Wei)