Lattice AI Research

Research focus

Distributed Systems & Hardware (2)Inference & Quantization (2)Computer Vision (1)Architecture Design (Transformers, SSMs, MoE) (1)

Frequent co-authors

Wenhui Ou (2)Zhuoyu Wu (2)Dongjun Wu (1)Freddy Ziyang Hong (1)

Papers (2)

Mar 1, 2026

Wenhui Ou +52w ago

FLICKER: A Fine-Grained Contribution-Aware Accelerator for Real-Time 3D Gaussian Splatting

Edge devices can now render 3D Gaussian Splats nearly 20x faster and with 27x better energy efficiency thanks to FLICKER, a contribution-aware hardware accelerator.

Wenhui Ou, Zhuoyu Wu, Yipu Zhang +3

Computer Vision Distributed Systems & Hardware Inference & Quantization

Wenhui Ou +32w ago

VIKIN: A Reconfigurable Accelerator for KANs and MLPs with Two-Stage Sparsity Support

KANs can run 1.25x faster and 4.87x more energy-efficiently than edge GPUs, thanks to a new reconfigurable accelerator that handles both KANs and MLPs.

Wenhui Ou, Zhuoyu Wu, Yipu Zhang +1

Architecture Design (Transformers, SSMs, MoE)Distributed Systems & Hardware Inference & Quantization

Search

Yipu Zhang

Publication activitypapers/week, last 8 weeks

Research focus

Frequent co-authors

Papers (2)