47 citations · 65 across the 13 of their papers we have counts for
13 papers
31.1 A 14.08-to-135.69Token/s ReRAM-on-Logic Stacked Outlier-Free Large-Language-Model Accelerator with Block-Clustered Weight-Compression and Adaptive Parallel-Speculative-Decoding
Pingcheng Dong, Yonghao Tan, Xuejiao Liu +13
This work presents a 55nm speculative decoding-based LLM accelerator with bumping-based face-to-face ReRAM-on-logic stacking technology. It features a local rotation unit for outli…
Expert Streaming: Accelerating Low-Batch MoE Inference via Multi-chiplet Architecture and Dynamic Expert Trajectory Scheduling
Songchen Ma, Hongyi Li, Weihao Zhang +8
Mixture-of-Experts is a promising approach for edge AI with low-batch inference. Yet, on-device deployments often face limited on-chip memory and severe workload imbalance; the pre…
A Deployment-Friendly Foundational Framework for Efficient Computational Pathology
Yu Cai, Cheng Jin, Zhengyu Zhang +25
Pathology foundation models (PFMs) generalize well across computational pathology tasks but remain costly for gigapixel whole-slide image analysis. Here, we present LitePath, a dep…
A 28nm 0.22μJ/token memory-compute-intensity-aware CNN-Transformer accelerator with hybrid-attention-based layer-fusion and cascaded pruning for semantic-segmentation
Pingcheng Dong, Yonghao Tan, Xuejiao Liu +14
This work presents a 28nm 13.93mm2 CNN-Transformer accelerator for semantic segmentation, achieving 3.86-to-10.91x energy reduction over previous designs. It features a hybrid atte…
Token Merging via Spatiotemporal Information Mining for Surgical Video Understanding
Xixi Jiang, Chen Yang, Dong Zhang +3
Vision Transformer models have shown impressive effectiveness in the surgical video understanding tasks through long-range dependency modeling. However, current methods suffer from…
APSQ: Additive Partial Sum Quantization with Algorithm-Hardware Co-Design
Yonghao Tan, Pingcheng Dong, Yongkun Wu +8
DNN accelerators, significantly advanced by model compression and specialized dataflow techniques, have marked considerable progress. However, the frequent access of high-precision…
Cyclic Contrastive Knowledge Transfer for Open-Vocabulary Object Detection
Chuhan Zhang, Chaoyang Zhu, Pingcheng Dong +2
In pursuit of detecting unstinted objects that extend beyond predefined categories, prior arts of open-vocabulary object detection (OVD) typically resort to pretrained vision-langu…
SR-LIO++: LiDAR-Inertial Odometry and Quantized Mapping with Caching-Aware Sweep Reconstruction
Zikang Yuan, Ruiye Ming, Chengwei Zhao +6
Addressing the inherent low acquisition frequency limitation of 3D LiDAR to achieve high-frequency output has become a critical research focus in the LiDAR-Inertial Odometry (LIO)…
Memory Efficient Transformer Adapter for Dense Predictions
Dong Zhang, Rui Yan, Pingcheng Dong +1
While current Vision Transformer (ViT) adapter methods have shown promising accuracy, their inference speed is implicitly hindered by inefficient memory access operations, e.g., st…
Genetic Quantization-Aware Approximation for Non-Linear Operations in Transformers
Pingcheng Dong, Yonghao Tan, Dong Zhang +11
Non-linear functions are prevalent in Transformers and their lightweight variants, incurring substantial and frequently underestimated hardware costs. Previous state-of-the-art wor…
Towards Customized Knowledge Distillation for Chip-Level Dense Image Predictions
Dong Zhang, Pingcheng Dong, Long Chen +1
It has been revealed that efficient dense image prediction (EDIP) models designed for AI chips, trained using the knowledge distillation (KD) framework, encounter two key challenge…
LLM-FP4: 4-Bit Floating-Point Quantized Transformers
Shih-yang Liu, Zechun Liu, Xijie Huang +2
We propose LLM-FP4 for quantizing both weights and activations in large language models (LLMs) down to 4-bit floating-point values, in a post-training manner. Existing post-trainin…
Quantization Variation: A New Perspective on Training Transformers with Low-Bit Precision
Xijie Huang, Zhiqiang Shen, Pingcheng Dong +1
Despite the outstanding performance of transformers in both language and vision tasks, the expanding computation and model size have increased the demand for efficient deployment.…