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.
arXiv:2608. 15127v1 Announce Type: cross Abstract: Agentic applications are shifting AI serving from isolated model inference to long-running workloads in which LLMs coordinate tools, environments, and persistent state.
arXiv:2607. 15516v1 Announce Type: cross Abstract: Production LLM deployments combine two cost-reduction primitives: prompt caching (a discounted rate for re-used token prefixes) and prompt compression (fewer tokens sent).
arXiv:2608. 14624v1 Announce Type: new Abstract: Multi-agent LLM systems have emerged as an important deployment paradigm for AI services, where each user request is decomposed into a sequence of specialized agents.
arXiv:2606. 09613v1 Announce Type: cross Abstract: Multi-turn LLM agents interleave model calls with external tool invocations, shifting serving from stateless request processing to stateful program execution.
arXiv:2607. 20495v1 Announce Type: new Abstract: Multi-agent systems decompose complex tasks into directed acyclic graphs (DAGs) of specialized agent executions, creating natural opportunities for caching intermediate results across queries.
arXiv:2606. 09916v1 Announce Type: cross Abstract: Multi-turn LLM agents fan short queries into long trajectories of tool calls, search results, and intermediate reasoning.
The paper investigates whether training Mixture-of-Experts (MoE) routers can improve memory‑bandwidth locality on consumer GPUs. Using a new zero‑surgery telemetry tool, the authors measure that a large Qwen3‑235B model is bottlenecked by disk‑based expert access, and that an LRU cache can serve a majority of requests. They pre‑register experiments training 137 M‑parameter MoE models with locality‑aware losses, finding that while cache misses can drop up to 60 % (99 % static‑pin hit rate), every configuration fails to meet a strict 1 % perplexity threshold, indicating a tight coupling between cache efficiency and model quality.
Multi-turn LLM agents interleave model calls with external tool invocations, shifting serving from stateless request processing to stateful program execution. Serving these workloads requires scheduling, KV-cache management, and routing policies that use program-level context, including turn dependencies, tool-induced gaps, and reusable KV state.
arXiv:2608. 19677v1 Announce Type: cross Abstract: Prefix caching avoids prefill only when a repeated request returns to a server that still holds the prefix KV.
arXiv:2606. 20474v1 Announce Type: cross Abstract: Context-heavy agents place unusual pressure on the key-value (KV) cache: long prefixes are reused across many short turns, while concurrency determines whether the serving system can keep GPUs utilized.
arXiv:2608. 06557v1 Announce Type: cross Abstract: The reasoning and agentic capabilities of large language models have expanded the range of applications they support, from short interactive exchanges to long, compute-heavy requests.
Serving a 235B-parameter Mixture-of-Experts (MoE) model on a single 8 GB GPU is bottlenecked not by compute but by memory bandwidth: decode must stream each token's active experts from whichever tier holds them, and on consumer hardware most experts sit on an SSD far slower than RAM. We quantify this bandwidth wall on Qwen3-235B (Q4_K_M, 134 GB): measured decode is 0.