# Serving Heterogeneous LoRA Adapters in Distributed LLM Inference Systems

arXiv: https://arxiv.org/abs/2511.22880 · HTML: https://mlsanity.com/paper/2511.22880/

## Key facts

- Cited by 5 papers in the map (data through 14 Oct 2026).
- Placed in Serving systems and request scheduling, one of 12 directions in Efficient LLM inference and serving.
- Builds on 28 of its 51 references that are in the map.
- First posted on arXiv 28 Nov 2025.

## Placed in

- Efficient LLM inference and serving / [Serving systems and request scheduling](https://mlsanity.com/inference/serving-systems-and-request-scheduling/)

## Builds on (most cited first, 28 of 28)

- [The Llama 3 Herd of Models](https://mlsanity.com/paper/2407.21783/) (2024)
- [LoRA: Low-Rank Adaptation of Large Language Models](https://mlsanity.com/paper/2106.09685/) (2021)
- [Prefix-Tuning: Optimizing Continuous Prompts for Generation](https://mlsanity.com/paper/2101.00190/) (2021)
- [The Power of Scale for Parameter-Efficient Prompt Tuning](https://mlsanity.com/paper/2104.08691/) (2021)
- [FlashAttention: Fast and Memory-Efficient Exact Attention with IO-Awareness](https://mlsanity.com/paper/2205.14135/) (2022)
- [OPT: Open Pre-trained Transformer Language Models](https://mlsanity.com/paper/2205.01068/) (2022)
- [QLoRA: Efficient Finetuning of Quantized LLMs](https://mlsanity.com/paper/2305.14314/) (2023)
- [Retrieval-Augmented Generation for Large Language Models: A Survey](https://mlsanity.com/paper/2312.10997/) (2023)
- [Code Llama: Open Foundation Models for Code](https://mlsanity.com/paper/2308.12950/) (2023)
- [LLM.int8(): 8-bit Matrix Multiplication for Transformers at Scale](https://mlsanity.com/paper/2208.07339/) (2022)
- [Few-Shot Parameter-Efficient Fine-Tuning is Better and Cheaper than In-Context Learning](https://mlsanity.com/paper/2205.05638/) (2022)
- [Splitwise: Efficient generative LLM inference using phase splitting](https://mlsanity.com/paper/2311.18677/) (2023)
- [Taming Throughput-Latency Tradeoff in LLM Inference with Sarathi-Serve](https://mlsanity.com/paper/2403.02310/) (2024)
- [Compacter: Efficient Low-Rank Hypercomplex Adapter Layers](https://mlsanity.com/paper/2106.04647/) (2021)
- [FlashInfer: Efficient and Customizable Attention Engine for LLM Inference Serving](https://mlsanity.com/paper/2501.01005/) (2025)
- [Parameter-Efficient Fine-Tuning Methods for Pretrained Language Models: A Critical Review and Assessment](https://mlsanity.com/paper/2312.12148/) (2023)
- [S-LoRA: Serving Thousands of Concurrent LoRA Adapters](https://mlsanity.com/paper/2311.03285/) (2023)
- [AdaMix: Mixture-of-Adaptations for Parameter-efficient Model Tuning](https://mlsanity.com/paper/2210.17451/) (2022)
- [MiniCache: KV Cache Compression in Depth Dimension for Large Language Models](https://mlsanity.com/paper/2405.14366/) (2024)
- [A Comprehensive Survey on Long Context Language Modeling](https://mlsanity.com/paper/2503.17407/) (2025)
- [DynamoLLM: Designing LLM Inference Clusters for Performance and Energy Efficiency](https://mlsanity.com/paper/2408.00741/) (2024)
- [Punica: Multi-Tenant LoRA Serving](https://mlsanity.com/paper/2310.18547/) (2023)
- [LoRA-Pro: Are Low-Rank Adapters Properly Optimized?](https://mlsanity.com/paper/2407.18242/) (2024)
- [Layer-Condensed KV Cache for Efficient Inference of Large Language Models](https://mlsanity.com/paper/2405.10637/) (2024)
- [Large Language Model Inference Acceleration: A Comprehensive Hardware Perspective](https://mlsanity.com/paper/2410.04466/) (2024)
- [BatchLLM: Optimizing Large Batched LLM Inference with Global Prefix Sharing and Throughput-oriented Token Batching](https://mlsanity.com/paper/2412.03594/) (2024)
- [Chameleon: Adaptive Caching and Scheduling for Many-Adapter LLM Inference Environments](https://mlsanity.com/paper/2411.17741/) (2024)
- [AerialDB: A Federated Peer-to-Peer Spatio-temporal Edge Datastore for Drone Fleets](https://mlsanity.com/paper/2508.07124/) (2025)

## Built on by (most cited first, 5 of 5)

- [MoLoRA: Composable Specialization via Per-Token Adapter Routing](https://mlsanity.com/paper/2603.15965/) (2026)
- [MinT: Managed Infrastructure for Training and Serving Millions of LLMs](https://mlsanity.com/paper/2605.13779/) (2026)
- [Data Driven Optimization of GPU efficiency for Distributed LLM-Adapter Serving](https://mlsanity.com/paper/2602.24044/) (2026)
- [Not All Ranks Are Equal: Budget-Aware LoRA Merging Across Tasks](https://mlsanity.com/paper/2609.22237/) (2026)
- [From Overloaded to Guaranteed: High-Throughput Multi-SLO Enforcement for LoRA-Assisted On-Premise LLM Deployment](https://mlsanity.com/paper/2610.04956/) (2026)

## Abstract (arXiv)

Low-Rank Adaptation (LoRA) has become the de facto method for parameter-efficient fine-tuning of large language models (LLMs), enabling rapid adaptation to diverse domains. In production, LoRA-based models are served at scale, creating multi-tenant environments with hundreds of adapters sharing a base model. However, state-of-the-art serving systems co-batch heterogeneous adapters without accounting for rank (size) variability, leading to severe performance skew, which ultimately requires adding more GPUs to satisfy service-level objectives (SLOs). Existing optimizations, focused on loading, caching, and kernel execution, ignore this heterogeneity, leaving GPU resources underutilized. We present LoRAServe, a workload-aware dynamic adapter placement and routing framework designed to tame rank diversity in LoRA serving. By dynamically rebalancing adapters across GPUs and leveraging GPU Direct RDMA for remote access, LoRAServe maximizes throughput and minimizes tail latency under real-world workload drift. Evaluations on production traces from Company X show that LoRAServe elicits up to 2× higher throughput, up to 9× lower TTFT, while using up to 50% fewer GPUs under SLO constraints compared to state-of-the-art systems.