The paper introduces GraphTok, a GPU-based tokenizer that executes the entire tokenization pipeline as a single CUDA graph, matching reference token IDs for models such as GPT‑2, Llama 3, Qwen 2.5, DeepSeek‑V3, and Gemma 3. GraphTok achieves significant speedups, encoding one million GPT‑2 tokens in 1.7–3.2 ms and outperforming the leading CPU tokenizer gigatoken by 4.7–6.0× on Llama 3. In a vLLM deployment with cached prefixes, it reduces mean time‑to‑first‑token by 2.1–2.4× and dramatically increases request throughput, sustaining 518 req/s of 121k‑token traffic versus 296 with gigatoken.
By Mohammad Siavashi, Pascal Schwinte, Ali Banaei, Gerald Q. Maguire Jr., Marco Chiesa, Dejan Kosti\'c
Diffusion language models (dLLMs) generate text by iteratively denoising a masked response and can commit multiple output positions per model invocation. Their bidirectional attention prevents exact autoregressive-style KV caching, since committing one position shifts the KV activations of all others.
The paper investigates how speculative decoding—where a draft model proposes multiple future tokens that are then verified by the target model—affects energy consumption in large language model serving under latency constraints. It finds that energy savings depend heavily on the acceptance rate of drafted tokens, which varies by model architecture, drafting scheme, and the type of text (e.g., conversation vs. code, English vs. Romansh). The study shows that while speculation can reduce energy per token at small batch sizes, at larger batches the verification cost dominates, potentially turning speculation into an energy loss, especially when acceptance rates are low.
By Bole Ma, Ayesha Afzal, Jan Eitzinger, Harald Koestler, Gerhard Wellein
arXiv:2609.37626v1 Announce Type: cross
Abstract: No single way of parallelizing attention serves large language models well under all loads. Low concurrency favors tensor parallelism, many independe...
By Chuan Liu, Shuoming Zhang, Zhicheng Li, Qianqi Sun, Ruiyuan Xu, Qiuchu Yu, Xiyu Shi, Huimin Cui, Jiacheng Zhao
arXiv:2607. 04206v1 Announce Type: cross Abstract: Diffusion language models (dLLMs) generate text by iteratively denoising a masked response and can commit multiple output positions per model invocation.
By Nitin Kedia, Saurabh Agarwal, Myungjin Lee, Aditya Akella
The paper investigates how prefix caching, a default optimization in open‑source LLM serving stacks, affects reproducibility when combined with weight quantization. Experiments on an eighty‑episode multi‑turn agentic tool‑use workload show that enabling the cache causes the agent’s trajectory to change in 36.2 % of episodes at 16‑bit precision and 75.0 % at 4‑bit precision, while disabling the cache yields perfectly reproducible runs. The study identifies specific cache‑related settings that drive run‑to‑run divergence and demonstrates that cached serving is deterministic only when the cache state is preserved, which is not the case in typical deployments.
By Aditi Patodiya
arXiv:2607. 21535v1 Announce Type: new Abstract: Speculative decoding accelerates autoregressive generation by having a cheap draft propose tokens that a target verifies in parallel.
By Alagappan Valliappan
The paper investigates the performance impact of using a remote KV cache tier for long‑context language model serving. In a disaggregated deployment of GLM‑5.3 on 8xB300 nodes, enabling the remote tier initially increased the 16k‑token p99 latency from 3.96 s to 17.3 s, and synchronous loading caused severe queue instability. By sharding the KV store across context‑parallel ranks, partitioning the remote tier, adding asynchronous loading, and fixing client bugs, the authors reduced the p99 latency from 83.32 s to 0.67 s and increased prefill throughput by 14–34 %.
whyItMatters":"The study demonstrates that careful sharding and asynchronous handling of remote KV caches can dramatically improve latency and throughput in large‑scale LLM serving systems."
By Muhammad Ibrahim
arXiv:2608.30310v1 Announce Type: cross
Abstract: Hybrid large language models interleave full-attention layers with linear-attention layers to reduce the cost of long-context inference. This structu...
By Yirui Liu, Ruoling Qi, Xuaner Wu, Penghang Liu, Jian Chen
Speculative decoding accelerates autoregressive generation by having a cheap draft propose tokens that a target verifies in parallel. Frontier models increasingly ship a built-in Multi-Token-Prediction (MTP/NEXTN) draft head under the assumption that the draft is negligibly cheap.
The paper introduces ACache, an affix-oriented cache reuse mechanism for Diffusion Large Language Models (DLLMs). ACache identifies a small set of critical affix tokens, called Anchor Tokens, and selectively recomputes their key-value states while reusing the rest of the affix cache. Experiments on Fast-dLLM and Nano-vLLM show that recomputing about 20% of affix tokens restores accuracy and can reduce recompute latency by up to 55.7% while improving throughput by up to 1.68×.
By Kaihua Liang, An Zhong, Xin Tan, Zafar Ayyub Qazi, Hong Xu, Jian Weng, Marco Canini
TokenRouter is a new serving system designed for token‑level routing of large language models (LLMs). It adopts a request‑centric programming model and launches a dedicated subserver for each LLM, using delayed‑batching schedulers whose hyperparameters are derived from a throughput model. Across various routing algorithms, workloads, and model pairs, TokenRouter delivers 2.01‑64.15× higher decoding throughput than existing systems, markedly improving the efficiency of token‑level LLM inference.
By Tianyu Fu, Tengxuan Liu, Ruoxi Wang, Yixin Dong, Yi Ge, Yichen You, Yu Wang