# KIS-S: A GPU-Aware Kubernetes Inference Simulator with RL-Based Auto-Scaling

arXiv: https://arxiv.org/abs/2507.07932 · HTML: https://mlsanity.com/paper/2507.07932/

## Key facts

- Cited by 9 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 10 Jul 2025.

## Built on by (most cited first, 9 of 9)

- [Adaptive GPU Resource Allocation for Multi-Agent Collaborative Reasoning in Serverless Environments](https://mlsanity.com/paper/2512.22149/) (2025)
- [When Does Deep RL Beat Calibrated Baselines? A Benchmark Study on Adaptive Resource Control](https://mlsanity.com/paper/2605.26418/) (2026)
- [Serverless GPU Architecture for Enterprise HR Analytics: A Production-Scale BDaaS Implementation](https://mlsanity.com/paper/2510.19689/) (2025)
- [NimbusGuard: A Novel Framework for Proactive Kubernetes Autoscaling Using Deep Q-Networks](https://mlsanity.com/paper/2604.11017/) (2026)
- [EvalStop: Using World Feedback to Detect and Correct Reward Overoptimization in Multi-Tenant RLHF Platforms](https://mlsanity.com/paper/2606.04145/) (2026)
- [Winning by Peeking: Unenforced Budgets and Test-Set Selection Inflate Short-Budget AutoML Comparisons](https://mlsanity.com/paper/2608.07303/) (2026)
- [Evaluating Kubernetes Performance for GenAI Inference: From Automatic Speech Recognition to LLM Summarization](https://mlsanity.com/paper/2602.04900/) (2026)
- [MAS-H2: A Hierarchical Multi-Agent System for Holistic Cloud-Native Autoscaling](https://mlsanity.com/paper/2603.07607/) (2026)
- [Learning to Remember: Attentive Reinforcement Learning for Edge Serverless Autoscaling](https://mlsanity.com/paper/2603.28790/) (2026)

## Abstract (arXiv)

Autoscaling GPU inference workloads in Kubernetes remains challenging due to the reactive and threshold-based nature of default mechanisms such as the Horizontal Pod Autoscaler (HPA), which struggle under dynamic and bursty traffic patterns and lack integration with GPU-level metrics. We present KIS-S, a unified framework that combines KISim, a GPU-aware Kubernetes Inference Simulator, with KIScaler, a Proximal Policy Optimization (PPO)-based autoscaler. KIScaler learns latency-aware and resource-efficient scaling policies entirely in simulation, and is directly deployed without retraining. Experiments across four traffic patterns show that KIScaler improves average reward by 75.2%, reduces P95 latency up to 6.7x over CPU baselines, and generalizes without retraining. Our work bridges the gap between reactive autoscaling and intelligent orchestration for scalable GPU-accelerated environments.