11 citations · 11 across the 10 of their papers we have counts for
12 papers
EAServe: Encode-Aware Disaggregated Serving for Multimodal Large Language Models
Kunxiong Zhu, Zhihao Shu, Hangyu Zheng +4
Disaggregating the two stages, Prefill and Decode, onto separate GPU pools is now a standard optimization for (text-only) LLM serving. However, multimodal LLMs (MLLMs), which add a…
Mask-Aware Execution for Efficient JEPA Training
Md Musfiqur Rahman Sanim, Zhihao Shu, Bahram Afsharmanesh +3
Joint Embedding Predictive Architectures (JEPAs) are becoming a core representation-learning primitive and a building block for latent world models across vision, video, audio, bra…
TierKV: Long-Context On-Device LLMs via Predictive Multi-Tier KV Caching
Zhihao Shu, Md Musfiqur Rahman Sanim, Jie Hu +4
Large language models (LLMs) are moving onto mobile devices for increasingly diverse workloads over text, images, video, and audio. These applications often require long contexts,…
FBLayout: Optimizing Memory Layout for Efficient LLM Finetuning on Mobile GPUs
Kahou Tam, Wei Niu, Yu Bao +3
Transformer-based models have enabled unprecedented capabilities across language, vision, and multimodal tasks. On-device fine-tuning of transformer models offers a privacy-preserv…
Splaxel: Efficient Distributed Training of 3D Gaussian Splatting for Large-scale Scene Reconstruction via Pixel-level Communication
Wenqi Jia, Zhewen Hu, Ying Huang +10
3D Gaussian Splatting (3DGS) enables high-fidelity and real-time 3D scene reconstruction, but scaling training to large-scale scenes requires optimizing hundreds of millions of Gau…
Gaussians on a Diet: High-Quality Memory-Bounded 3D Gaussian Splatting Training
Yangming Zhang, Jian Xu, Chaojian Li +7
3D Gaussian Splatting (3DGS) has revolutionized novel view synthesis with high-quality rendering through continuous aggregations of millions of 3D Gaussian primitives. However, it…
FlashMem: Supporting Modern DNN Workloads on Mobile with GPU Memory Hierarchy Optimizations
Zhihao Shu, Md Musfiqur Rahman Sanim, Hangyu Zheng +4
The increasing size and complexity of modern deep neural networks (DNNs) pose significant challenges for on-device inference on mobile GPUs, with limited memory and computational r…
AdaCM: On Understanding Extremely Long-Term Video with Adaptive Cross-Modality Memory Reduction
Yuanbin Man, Ying Huang, Chengming Zhang +3
The advancements in large language models (LLMs) have propelled the improvement of video understanding tasks by incorporating LLMs with visual models. However, most existing LLM-ba…
GaussianSpa: An "Optimizing-Sparsifying" Simplification Framework for Compact and High-Quality 3D Gaussian Splatting
Yangming Zhang, Wenqi Jia, Wei Niu +1
3D Gaussian Splatting (3DGS) has emerged as a mainstream for novel view synthesis, leveraging continuous aggregations of Gaussian functions to model scene geometry. However, 3DGS s…
NeurLZ: An Online Neural Learning-Based Method to Enhance Scientific Lossy Compression
Wenqi Jia, Zhewen Hu, Youyuan Liu +10
Large-scale scientific simulations generate massive datasets, posing challenges for storage and I/O. Traditional lossy compression struggles to advance more in balancing compressio…
SmartMem: Layout Transformation Elimination and Adaptation for Efficient DNN Execution on Mobile
Wei Niu, Md Musfiqur Rahman Sanim, Zhihao Shu +5
This work is motivated by recent developments in Deep Neural Networks, particularly the Transformer architectures underlying applications such as ChatGPT, and the need for performi…
GWLZ: A Group-wise Learning-based Lossy Compression Framework for Scientific Data
Wenqi Jia, Sian Jin, Jinzhen Wang +3
The rapid expansion of computational capabilities and the ever-growing scale of modern HPC systems present formidable challenges in managing exascale scientific data. Faced with su…