Related papers: Swarm: Co-Activation Aware KVCache Offloading Acro…

Memory-augmented Large Language Models (LLMs) have demonstrated remarkable capability for complex and long-horizon embodied planning. By keeping track of past experiences and environmental states, memory enables LLMs to maintain a global…

Traditional KV cache eviction strategies, which discard less critical KV-pairs based on attention scores, often degrade generation quality, causing context loss or hallucinations. Recent efforts shift toward KV merging, merging eviction…

Vision-Language Models (VLMs) have demonstrated impressive performance across a versatile set of tasks. A key challenge in accelerating VLMs is storing and accessing the large Key-Value (KV) cache that encodes long visual contexts, such as…

Cache-assisted ultra-dense mobile edge computing (MEC) networks are a promising solution for meeting the increasing demands of numerous Internet-of-Things mobile devices (IMDs). To address the complex interferences caused by small base…

In this paper we introduce our vision of a Cognitive Computing Continuum to address the changing IT service provisioning towards a distributed, opportunistic, self-managed collaboration between heterogeneous devices outside the traditional…

We present Keigo, a concurrency- and workload-aware storage middleware that enhances the performance of log-structured merge key-value stores (LSM KVS) when they are deployed on a hierarchy of storage devices. The key observation behind…

Sparse matrix-vector multiplication (SpMV) is crucial in computational science, engineering, and machine learning. Despite substantial efforts to improve SpMV performance on GPUs through various techniques, issues related to data locality,…

Vehicular ad hoc networks (VANETs) allow vehicles to exchange warning messages with each other. These specific kinds of networks help reduce hazardous traffic situations and improve safety, which are two of the main objectives in developing…

Key-Value cache (\texttt{KV} \texttt{cache}) compression has emerged as a promising technique to optimize Large Language Model (LLM) serving. It primarily decreases the memory consumption of \texttt{KV} \texttt{cache} to reduce the…

Recent large language models (LLMs) face increasing inference latency as input context length and model size continue to grow. In particular, the retrieval-augmented generation (RAG) technique, which enhances LLM responses by incorporating…

While large language models (LLMs) excel at handling long-context sequences, they require substantial prefill computation and key-value (KV) cache, which can heavily burden computational efficiency and memory usage in both prefill and…

Coflow has emerged as a fundamental application-layer abstraction in distributed systems, representing communication dependencies and enabling collaborative management of related flows to enhance job completion efficiency. To meet the…

Serving LLMs with a cluster of GPUs is common nowadays, where the serving system must meet strict latency SLOs required by applications. However, the stateful nature of LLM serving requires maintaining huge states (i.e., KVCache) in limited…

Video Large Language Models (Video-LLMs) have demonstrated significant potential in the areas of video captioning, search, and summarization. However, current Video-LLMs still face challenges with long real-world videos. Recent methods have…

Multimodal large language models suffer from substantial inference overhead since multimodal KV Cache grows proportionally with the visual input length. Existing multimodal KV Cache compression methods mostly rely on attention score to…

Large language models (LLMs) are typically served from clusters of GPUs/NPUs that consist of large number of devices. Unfortunately, communication between these devices incurs significant overhead, increasing the inference latency and cost…

We demonstrate that unstructured sparsity significantly improves KV cache compression for LLMs, enabling sparsity levels up to 70% without compromising accuracy or requiring fine-tuning. We conduct a systematic exploration of pruning…

With the rapid growth of dynamic vision sensor (DVS) data, constructing a low-energy, efficient data retrieval system has become an urgent task. Hash learning is one of the most important retrieval technologies which can keep the distance…

A growing trend has emerged in designing high-quality Small Language Models (SLMs) with a few million parameters. This trend is driven by the increasing concerns over cloud costs, privacy, and latency. Considering that full parameter…

The increasing adoption of large language models (LLMs) with extended context windows necessitates efficient Key-Value Cache (KVC) management to optimize inference performance. Inference workloads like Retrieval-Augmented Generation (RAG)…