Related papers: Conduit: Programmer-Transparent Near-Data Processi…
Multi Scale Deformable Attention (MSDAttn) has become a fundamental component in various vision tasks due to its effective multi scale grid sampling (MSGS). However, its reliance on random sampling results in highly irregular memory access…
Nonequispaced discrete Fourier transformation (NDFT) is widely applied in all aspects of computational science and engineering. The computational efficiency and accuracy of NDFT has always been a critical issue in hindering its…
This paper presents a novel sensitivity-based distributed programming (SBDP) approach for non-convex, large-scale nonlinear programs (NLP). The algorithm relies on first-order sensitivities to cooperatively solve the central NLP in a…
Sliding window convolutional networks (ConvNets) have become a popular approach to computer vision problems such as image segmentation, and object detection and localization. Here we consider the problem of inference, the application of a…
Data stream processing systems (DSPSs) enable users to express and run stream applications to continuously process data streams. To achieve real-time data analytics, recent researches keep focusing on optimizing the system latency and…
SmartNIC Data Processing Units (DPUs) offer a promising solution for saving high-end CPU resources by offloading tasks to programmable cores near the network interface. In this work, we explore the feasibility of SmartNIC DPUs in supporting…
With the widespread use of deep neural networks(DNNs) in intelligent systems, DNN accelerators with high performance and energy efficiency are greatly demanded. As one of the feasible processing-in-memory(PIM) architectures,…
The rapid growth of deep learning has driven exponential increases in model parameters and computational demands. NVIDIA GPUs and their CUDA-based software ecosystem provide robust support for parallel computing, significantly alleviating…
Compute-Near-Memory (CNM) systems offer a promising approach to mitigate the von Neumann bottleneck by bringing computational units closer to data. However, optimizing for these architectures remains challenging due to their unique hardware…
LiDAR-based perception is critical for autonomous driving due to its robustness to poor lighting and visibility conditions. Yet, current models operate under the closed-set assumption and often fail to recognize unexpected…
Eliminating duplicate data in primary storage of clouds increases the cost-efficiency of cloud service providers as well as reduces the cost of users for using cloud services. Existing primary deduplication techniques either use inline…
Near-data accelerators (NDAs) that are integrated with main memory have the potential for significant power and performance benefits. Fully realizing these benefits requires the large available memory capacity to be shared between the host…
The constant growth of DNNs makes them challenging to implement and run efficiently on traditional compute-centric architectures. Some accelerators have attempted to add more compute units and on-chip buffers to solve the memory wall…
Deploying deep neural networks (DNNs) on resource-constrained mobile devices presents significant challenges, particularly in achieving real-time performance while simultaneously coping with limited computational resources and battery life.…
Efficiently scaling deep neural networks across GPU clusters requires navigating complex trade-offs between computational throughput, memory utilization, and synchronization overhead. This paper presents a unified empirical evaluation of…
Token reduction accelerates Multimodal Large Language Models (MLLMs) by reducing excessive tokens, but overlooks structural redundancy differences, where critical and redundant modules process identical token loads. For fine-grained…
The growth rate of the GPU memory capacity has not been able to keep up with that of the size of large language models (LLMs), hindering the model training process. In particular, activations -- the intermediate tensors produced during…
Sparse training is one of the promising techniques to reduce the computational cost of DNNs while retaining high accuracy. In particular, N:M fine-grained structured sparsity, where only N out of consecutive M elements can be nonzero, has…
Networking is considered a datacenter tax, and hyperscalers push hard to provide high-performance networking with minimal resource expenditure. To keep up with the ever-increasing network rates, many CPU cycles are spent on the networking…
High-performance computing systems are moving towards 2.5D and 3D memory hierarchies, based on High Bandwidth Memory (HBM) and Hybrid Memory Cube (HMC) to mitigate the main memory bottlenecks. This trend is also creating new opportunities…