6 papers
Optimal Design of Shaped Pupil Lyot Coronagraph for China's Large Segmented Telescope
Kaining Yang, He Sun, Bin B. Ren +7
China's next-generation large telescopes, such as the Expandable Aperture Segmented Telescope (EAST) and the Large Optical/Infrared Telescope (LOT), are being planned as flagship f…
Rethinking Heterogeneous System Disaggregation for Subquadratic Attention
Arya Tschand, Yaosheng Fu, Vikram Sharma Mailthody +6
Frontier language models are more aggressively using subquadratic attention to reduce the memory footprint and compute requirements during inference while still delivering frontier…
Automated Synthesis of Heterogeneous, Hierarchical, Scoped Coherence Protocols
Fletch Rydell, An Qi Zhang, Nicolai Oswald +3
Processor design is converging on a new model of cache-coherent shared memory characterized by heterogeneity, hierarchy, and scopes. Protocols like CXL or AMBA CHI are used as glob…
Nemotron-Labs-Diffusion: A Tri-Mode Language Model Unifying Autoregressive, Diffusion, and Self-Speculation Decoding
Yonggan Fu, Lexington Whalen, Abhinav Garg +23
We introduce Nemotron-Labs-Diffusion, a tri-mode language model (LM) that unifies AR, diffusion, and self-speculation decoding within a single architecture. Trained with a joint AR…
-dLLM: Accelerating Diffusion Large Language Models via Spatio-Temporal Redundancy Reduction
Zhenbang Du, Kejing Xia, Xinrui Zhong +6
Diffusion Large Language Models (dLLMs) have emerged as a promising alternative to autoregressive generation by enabling parallel token prediction. However, practical dLLM decoding…
Efficient MoE Serving in the Memory-Bound Regime: Balance Activated Experts, Not Tokens
Yanpeng Yu, Haiyue Ma, Krish Agarwal +10
Expert Parallelism (EP) permits Mixture of Experts (MoE) models to scale beyond a single GPU. To address load imbalance across GPUs in EP, existing approaches aim to balance the nu…