Related papers: A Host-SSD Collaborative Write Accelerator for LSM…

Prefix KV caching has become a key mechanism in LLM serving: it reduces time to first token (TTFT) by avoiding redundant computation across requests that share a prefix (i.e., the system prompt). However, the accumulated KV cache is often…

Solid state disks (SSDs) have advanced to outperform traditional hard drives significantly in both random reads and writes. However, heavy random writes trigger fre- quent garbage collection and decrease the performance of SSDs. In an SSD…

We present MemEC, an erasure-coding-based in-memory key-value (KV) store that achieves high availability and fast recovery while keeping low data redundancy across storage servers. MemEC is specifically designed for workloads dominated by…

To overcome the well-known memory bottleneck of AI chips, 3D stacked architectures that employ advanced packaging technology with high-density through-silicon vias (TSVs) pins have proven to be a promising solution. The 3D-stacked AI chip…

The Key-Value (KV) cache is the primary memory bottleneck in long-context Large Language Models, yet it is typically treated as an opaque numerical tensor. In this work, we propose \textbf{STA-Attention}, a framework that utilizes Top-K…

Storing digital information, ensuring the accuracy, steady and uninterrupted access to the data are considered as fundamental challenges in enterprise-class organizations and companies. In recent years, new types of storage systems such as…

This paper introduces NVCache, an approach that uses a non-volatile main memory (NVMM) as a write cache to improve the write performance of legacy applications. We compare NVCache against file systems tailored for NVMM (Ext4-DAX and NOVA)…

Recent announcements have shown the viability of end-to-end open-source (OS) Linux-capable RISC-V systems on chip (SoCs). However, practical application and software development platforms require efficient non-volatile storage, which is not…

The energy efficiency of analog computing-in-memory (ACIM) accelerator for recurrent neural networks, particularly long short-term memory (LSTM) network, is limited by the high proportion of nonlinear (NL) operations typically executed…

Elasticity is one important feature in modern cloud computing systems and can result in computation failure or significantly increase computing time. Such elasticity means that virtual machines over the cloud can be preempted under a short…

Large language models (LLMs) can now handle longer sequences of tokens, enabling complex tasks like book understanding and generating lengthy novels. However, the key-value (KV) cache required for LLMs consumes substantial memory as context…

Supporting long-context LLMs is challenging due to the substantial memory demands of the key-value (KV) cache. Existing offloading systems store the full cache in host memory and selectively fetch critical entries during decoding, but this…

Quad-level cell (QLC) flash offers significant benefits in cost and capacity, but its limited reliability leads to frequent read retries, which severely degrade read performance. A common strategy in high-density flash storage is to program…

In recent years, high interest in using Virtual Machines (VMs) in data centers and Cloud computing has significantly increased the demand for high-performance data storage systems. Recent studies suggest using SSDs as a caching layer for…

In the modern era of multicore processors, utilizing cores is a tedious job. Synchronization and communication among processors involve high cost. Software transaction memory systems (STMs) addresses this issues and provide better…

Vision-language models (VLMs) show remarkable performance in multimodal tasks. However, excessively long multimodal inputs lead to oversized Key-Value (KV) caches, resulting in significant memory consumption and I/O bottlenecks. Previous KV…

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…

The cloud infrastructure motivates disaggregation of monolithic data stores into components that are assembled together based on an application's workload. This study investigates disaggregation of an LSM-tree key-value store into…

This paper presents new alternatives to the well-known Bloom filter data structure. The Bloom filter, a compact data structure supporting set insertion and membership queries, has found wide application in databases, storage systems, and…

The growing memory footprint of the Key-Value (KV) cache poses a severe scalability bottleneck for long-context Large Language Model (LLM) inference. While KV cache eviction has emerged as an effective solution by discarding less critical…