Related papers: DualScale: Energy-Efficient Disaggregated LLM Serv…

In this chapter we focus on slack reclamation and propose a new slack reclamation technique, Multiple Frequency Selection DVFS (MFS-DVFS). The key idea is to execute each task with a linear combination of more than one frequency such that…

Large language models (LLMs) excel across diverse tasks but face significant deployment challenges due to high inference costs. LLM inference comprises prefill (compute-bound) and decode (memory-bound) stages, with decode dominating latency…

The continuous growth of big data applications with high computational and scalability demands has resulted in increasing popularity of cloud computing. Optimizing the performance and power consumption of cloud resources is therefore…

Machine learning (ML) computations commonly execute on expensive specialized hardware, such as GPUs and TPUs, which provide high FLOPs and performance-per-watt. For cost efficiency, it is essential to keep these accelerators highly…

Large Language Models (LLMs) have become the new foundation for many applications, reshaping human society like a storm. Disaggregated inference, which separates prefill and decode stages, is a promising approach to improving hardware…

Serving disaggregated large language models (LLMs) over tens of thousands of xPU devices (GPUs or NPUs) with reliable performance faces multiple challenges. 1) Ignoring the diversity (various prefixes and tidal requests), treating all the…

Large language models (LLMs) have transformed natural language processing but face critical deployment challenges in device-edge systems due to resource limitations and communication overhead. To address these issues, collaborative…

The pressing demands of improving energy efficiency for high performance scientific computing have motivated a large body of software-controlled hard- ware solutions using Dynamic Voltage and Frequency Scaling (DVFS) that strategically…

Due to limited resources on edge and different characteristics of deep neural network (DNN) models, it is a big challenge to optimize DNN inference performance in terms of energy consumption and end-to-end latency on edge devices. In…

Large multimodal models (LMMs) demonstrate impressive capabilities in understanding images, videos, and audio beyond text. However, efficiently serving LMMs in production environments poses significant challenges due to their complex…

Deploying multiple models within shared GPU clusters is a key strategy to improve resource efficiency in large language model (LLM) serving. Existing multi-LLM serving systems improve GPU utilization at the cost of degraded inference…

Power consumption in data centers has been growing significantly in recent years. To reduce power, servers are being equipped with increasingly sophisticated power management mechanisms. Different mechanisms offer dramatically different…

Prefill-decode (PD) disaggregation has become the standard architecture for large-scale LLM serving, but in practice its deployment boundary is still determined by KVCache transfer. In conventional dense-attention models, prefill generates…

The performance of multi-turn, agentic LLM inference is increasingly dominated by KV-Cache storage I/O rather than computation. In prevalent disaggregated architectures, loading the massive KV-Cache from external storage creates a…

In multi-model LLM serving, decode execution remains inefficient due to model-specific resource partitioning: since cross-model batching is not possible, memory-bound decoding often suffers from severe GPU underutilization, especially under…

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…

With the continuous improvement of on-chip integrated voltage regulators (IVRs) and fast, adaptive frequency control, dynamic voltage-frequency scaling (DVFS) transition times have shrunk from the microsecond to the nanosecond regime,…

LoRA enables efficient customization of LLMs and is widely used in multi-tenant and multi-task serving. However, emerging model architectures such as MoE significantly increase LoRA memory cost, making existing coupled LoRA serving designs…

Distributed Stream Processing (DSP) systems are capable of processing large streams of unbounded data, offering high throughput and low latencies. To maintain a stable Quality of Service (QoS), these systems require a sufficient allocation…

Diffusion-based large language models (dLLMs) have recently gained significant attention for their exceptional performance and inherent potential for parallel decoding. Existing frameworks further enhance its inference efficiency by…