Related papers: Sandwich: Joint Configuration Search and Hot-Switc…

Large language models (LLMs) are becoming increasingly capable at small parameter scales. At the same time, conventional cloud-centric deployment introduces challenges around data privacy, latency, and cost that are acute in operational…

Memory-bound algorithms show complex performance and energy consumption behavior on multicore processors. We choose the lattice-Boltzmann method (LBM) on an Intel Sandy Bridge cluster as a prototype scenario to investigate if and how…

Large language models (LLMs) have surged in popularity and are extensively used in commercial applications, where the efficiency of model serving is crucial for the user experience. Most current research focuses on optimizing individual…

With the advent of exascale computing, effective load balancing in massively parallel software applications is critically important for leveraging the full potential of high performance computing systems. Load balancing is the distribution…

Large language models (LLMs) are increasingly deployed on customer devices. To support them, current devices are adopting SoCs (System on Chip) with NPUs (Neural Processing Unit) installed. Although high performance is expected, LLM…

Each LLM serving request goes through two phases. The first is prefill which processes the entire input prompt and produces the first output token and the second is decode which generates the rest of output tokens, one-at-a-time. Prefill…

The significant resource demands in LLM serving prompts production clusters to fully utilize heterogeneous hardware by partitioning LLM models across a mix of high-end and low-end GPUs. However, existing parallelization approaches often…

Traditional cluster designs were originally server-centric, and have evolved recently to support hardware acceleration and storage disaggregation. In applications that leverage acceleration, the server CPU performs the role of orchestrating…

Large language models (LLMs) have shown great potential in natural language processing and content generation. However, current LLMs heavily rely on cloud computing, leading to prolonged latency, high bandwidth cost, and privacy concerns.…

Current large language model (LLM) serving systems, primarily designed for text completion, are neither efficient nor adaptable for increasingly complex LLM applications due to their inflexible design. We propose a new LLM serving system…

Emerging large language model (LLM) applications involve diverse reasoning strategies and agentic workflows, straining the capabilities of existing serving systems built on a monolithic token generation loop. This paper introduces Pie, a…

LAPS identifies and disaggregates requests with different prompt lengths in LLM serving to reduce TTFT latency. While recent systems have decoupled the prefill and decode stages to improve throughput, they still rely on unified scheduling…

Tensor parallelism (TP) in large-scale LLM inference and training introduces frequent collective operations that dominate inter-GPU communication. While in-switch computing, exemplified by NVLink SHARP (NVLS), accelerates collective…

Serving numerous users and requests concurrently requires good fairness in Large Language Models (LLMs) serving system. This ensures that, at the same cost, the system can meet the Service Level Objectives (SLOs) of more users , such as…

Modern LLM serving systems must sustain high throughput while meeting strict latency SLOs across two distinct inference phases: compute-intensive prefill and memory-bound decode phases. Existing approaches either (1) aggregate both phases…

Large Language Models (LLMs) have resulted in a surging demand for planet-scale serving systems, where tens of thousands of GPUs continuously serve hundreds of millions of users. Consequently, throughput has emerged as a key metric that…

On-device inference for Large Language Models (LLMs), driven by increasing privacy concerns and advancements of mobile-sized models, has gained significant interest. However, even mobile-sized LLMs (e.g., Gemma-2B) encounter unacceptably…

Large Language Models (LLMs) such as GPT-4 and Llama have shown remarkable capabilities in a variety of software engineering tasks. Despite the advancements, their practical deployment faces challenges, including high financial costs, long…

Deploying large language models (LLMs) on edge devices is crucial for delivering fast responses and ensuring data privacy. However, the limited storage, weight, and power of edge devices make it difficult to deploy LLM-powered applications.…

With emerging storage-class memory (SCM) nearing commercialization, there is evidence that it will deliver the much-anticipated high density and access latencies within only a few factors of DRAM. Nevertheless, the latency-sensitive nature…