The paper systematically studies decode‑phase energy consumption of open‑source large language models using different attention architectures—Multi‑Head Attention (MHA), Grouped Query Attention (GQA), and GQA with Sliding Window Attention (SWA). It evaluates four models across varying context lengths, batch sizes, and generation workloads, measuring GPU energy via NVIDIA counters. Findings show that the attention mechanism is the main driver of how energy scales with context length, with MHA models growing steeply, GQA models growing less, and GQA+SWA remaining nearly constant; model size mainly sets absolute energy use, while batching can cut energy per token and latency by up to 87%.
By Molka Chkir, Syed Muhammad Danish, Jos H\"oll, Arghavan Asad
The paper introduces a phase‑decoupled, model‑calibrated power controller for disaggregated large‑language‑model (LLM) serving, addressing the mismatch between GPU power settings and the distinct hardware regimes of prefill and decode stages. By calibrating separate power caps for each lane based on measured throughput‑latency cliffs, the authors achieve a 20.4% increase in tokens per joule with only a 3.5% rise in mean end‑to‑end latency on an 8‑node B200 cluster, outperforming NVIDIA’s Max‑Q profile. The approach also demonstrates consistent meeting of ITL‑p99 service‑level objectives across multiple MoE models and yields a 32.3% electricity savings over a sustained three‑day run.
By Jae Gon Kim, Donghoon Yoo, Hanyul Ryu, Sungho Ha, Juyeon Lee, Soojung Ryu
The paper evaluates NVIDIA’s Max‑Q inference profile on a disaggregated B200 GPU system for large language model (LLM) serving, finding modest gains (+8.6% tokens/J) but increased latency (+5.2%). It proposes a phase‑decoupled, model‑calibrated power controller that sets a latency‑guaranteed SM‑clock window for prefill and a calibrated power cap for decode, achieving a 20.4% tokens/J improvement with only a 3.5% latency increase on an 8‑node Qwen3‑Coder‑480B deployment. The approach outperforms vendor profiles on both energy and latency, and demonstrates significant long‑term electricity savings in MoE‑based serving.
arXiv:2608.28044v1 Announce Type: cross
Abstract: Large language model (LLM) inference serving is priced by tokens, but GPU energy is consumed over inference windows. This accounting mismatch makes t...
By Prabhu Vellaisamy, Vanessa Lam, Shawn Blanton, John Paul Shen
The paper reports a reproducible GPU power benchmark for 18 open‑source LLMs (0.5B–7B parameters) run on a single consumer RTX 4060ti GPU using the Ollama inference engine. Energy metrics such as mean/peak power, total energy per prompt, energy per output token, and throughput were measured, revealing that model architecture and quantization strategy, rather than parameter count alone, drive energy efficiency. The most efficient models were qwen2.5:0.5b and tinyllama:1.1b, while the 7B‑Mistral model consumed up to 8.6× more energy per token, and qwen3.5:0.8b(on) showed unusually high per‑prompt energy due to extended internal reasoning.
By Philipp M. Z\"ahl, Elja Dalipaj, Anika Hennig, Timon Bayer
arXiv:2604. 26968v2 Announce Type: replace-cross Abstract: Key-value (KV) cache memory management is the primary bottleneck limiting throughput and cost-efficiency in large-scale GPU inference serving.
By Sanjeev Rao Ganjihal
AMMA is a multi‑chiplet, memory‑centric architecture designed to reduce latency in large‑context attention for language models. It replaces GPU compute dies with HBM‑PNM cubes, doubling memory bandwidth and employing a logic‑die microarchitecture, two‑level hybrid parallelism, and a reordered collective flow to minimize communication overhead. Design‑space exploration and evaluations show AMMA achieves 15.5× lower attention latency and 6.9× lower energy consumption compared with the NVIDIA H100.
By Zhongkai Yu, Haotian Ye, Chenyang Zhou, Ohm Rishabh Venkatachalam, Zaifeng Pan, Zhengding Hu, Junsung Kim, Won Woo Ro, Po-An Tsai, Shuyi Pei, Yangwook Kang, Yufei Ding
arXiv:2607. 02043v1 Announce Type: cross Abstract: Disaggregated LLM serving runs prefill and decode on separate GPU pools to keep the two phases from interfering.
By Shrikara Arun, Anjaly Parayil, Srikant Bharadwaj, Renee St. Amant, Victor R\"uhle
arXiv:2507. 07247v2 Announce Type: replace-cross Abstract: As large language models (LLMs) and visual language models (VLMs) grow in scale and application, attention mechanisms have become a central computational bottleneck due to their high memory and time complexity.
By Zhengyu Tian, Anantha Padmanaban Krishna Kumar, Hemant Krishnakumar, Reza Rawassizadeh
DualPace is an LLM serving runtime that decouples scheduling between a GPU engine and an AMX‑enabled CPU engine, allowing the CPU to independently batch and advance decodes while still cooperating with the GPU for dense stages. This design expands key‑value cache capacity into host memory and reduces bursty CPU progress, leading to significant throughput gains—up to 43.2 % over NEO and 90.6 % over vLLM—without increasing latency. The system also maintains low time‑to‑first‑token and supports higher request rates at comparable tail latency.
By YuXuan Lin, KeQiang Wu, Kingsum Chow
arXiv:2606. 17104v1 Announce Type: cross Abstract: As large language models (LLMs) are increasingly deployed in latency- and cost-sensitive settings, inference efficiency has become a central systems challenge.
By Shun Usami, Venkatram Vishwanath, E. Wes Bethel
arXiv:2606. 02964v1 Announce Type: cross Abstract: Large Language Model (LLM) inference relies on key-value (KV) caches to avoid redundant attention computation.
By Chunan Shi, Yilei Chen, Yilin Chen, Xupeng Miao, Bin Cui