Related papers: Analysis and Evaluation of Using Microsecond-Laten…

The key-value (KV) cache has become the dominant contributor to memory consumption in large language model (LLM) inference. Although offloading KVCache from GPU high-bandwidth memory (HBM) to CPU DRAM alleviates device memory pressure, DRAM…

In long-context Large Language Model (LLM) inference, the Time-To-First-Token (TTFT) latency incurred by the prefill stage has become the foremost bottleneck limiting interactive performance and deployment cost. KV Cache reuse offers a…

Recent large language models (LLMs) with enormous model sizes use many GPUs to meet memory capacity requirements incurring substantial costs for token generation. To provide cost-effective LLM inference with relaxed latency constraints,…

The impressive capabilities of Large Language Models (LLMs) come at the cost of substantial computational resources during deployment. While KV Cache can significantly reduce recomputation during inference, it also introduces additional…

Transformers have emerged as the underpinning architecture for Large Language Models (LLMs). In generative language models, the inference process involves two primary phases: prompt processing and token generation. Token generation, which…

Memory latencies and bandwidth are major factors, limiting system performance and scalability. Modern CPUs aim at hiding latencies by employing large caches, out-of-order execution, or complex hardware prefetchers. However, software-based…

Transformer-based large language models (LLMs) use the key-value (KV) cache to significantly accelerate inference by storing the key and value embeddings of past tokens. However, this cache consumes significant GPU memory. In this work, we…

Log-structured merge tree (LSM-tree) based key-value stores are widely employed in large-scale storage systems. In the compaction of the key-value store, SSTables are merged with overlapping key ranges and sorted for data queries. This,…

The increasing size of the Key-Value (KV) cache during the Large Language Models long-context inference is the main obstacle for its balance between the deployment cost and task accuracy. To reduce the KV cache size in such scenarios, most…

Owing to the huge success of generative artificial intelligence (AI), large language models (LLMs) have emerged as a core subclass, underpinning applications such as question answering, text generation, and code completion. While…

Large Language Models capable of handling extended contexts are in high demand, yet their inference remains challenging due to substantial Key-Value cache size and high memory bandwidth requirements. Previous research has demonstrated that…

The efficiency of Large Language Model~(LLM) inference is often constrained by substantial memory bandwidth and capacity demands. Existing techniques, such as pruning, quantization, and mixture of experts/depth, reduce memory capacity…

The rise of million-token, agent-based applications has placed unprecedented demands on large language model (LLM) inference services. The long-running nature of these tasks increases their susceptibility to hardware and software faults,…

Several works have developed eviction policies to remove key-value (KV) pairs from the KV cache for more efficient inference. The focus has been on compressing the KV cache after the input prompt has been processed for faster token…

In this paper we analyze the influence that lower layers (file system, OS, SSD) have on HDFS' ability to extract maximum performance from SSDs on the read path. We uncover and analyze three surprising performance slowdowns induced by lower…

Deploying large language models (LLMs) on edge devices enables personalized agents with strong privacy and low cost. However, with tens to hundreds of billions of parameters, single-batch autoregressive inference suffers from extremely low…

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…

The increasing memory demand of the Key-Value (KV) cache poses a significant bottleneck for Large Language Models (LLMs) in long-context applications. Existing low-rank KV compression methods reduce this footprint by modifying model…

In this study, we introduce adaptive KV cache compression, a plug-and-play method that reduces the memory footprint of generative inference for Large Language Models (LLMs). Different from the conventional KV cache that retains key and…

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…