# Non-Monotonic Latency in Apple MPS Decoding: KV Cache Interactions and Execution Regimes

arXiv: https://arxiv.org/abs/2605.08913 · HTML: https://mlsanity.com/paper/2605.08913/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in KV cache optimization, one of 12 directions in Efficient LLM inference and serving.
- Builds on 12 of its 22 references that are in the map.
- First posted on arXiv 9 May 2026.

## Placed in

- Efficient LLM inference and serving / [KV cache optimization](https://mlsanity.com/inference/kv-cache-optimization/)

## Builds on (most cited first, 12 of 12)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [PyTorch: An Imperative Style, High-Performance Deep Learning Library](https://mlsanity.com/paper/1912.01703/) (2019)
- [Efficient Memory Management for Large Language Model Serving with PagedAttention](https://mlsanity.com/paper/2309.06180/) (2023)
- [DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lighter](https://mlsanity.com/paper/1910.01108/) (2019)
- [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)
- [BLOOM: A 176B-Parameter Open-Access Multilingual Language Model](https://mlsanity.com/paper/2211.05100/) (2022)
- [Efficiently Scaling Transformer Inference](https://mlsanity.com/paper/2211.05102/) (2022)
- [LLM in a flash: Efficient Large Language Model Inference with Limited Memory](https://mlsanity.com/paper/2312.11514/) (2023)
- [Dissecting the NVidia Turing T4 GPU via Microbenchmarking](https://mlsanity.com/paper/1903.07486/) (2019)
- [KV Prediction for Improved Time to First Token](https://mlsanity.com/paper/2410.08391/) (2024)
- [Training-Free Exponential Context Extension via Cascading KV Cache](https://mlsanity.com/paper/2406.17808/) (2024)

## Abstract (arXiv)

Autoregressive inference is typically assumed to scale predictably with decoding length, with latency increasing smoothly as generated sequence length grows. In this work, we identify unexpected non-monotonic latency behavior in the Apple MPS backend, where latency changes abruptly across nearby decoding configurations during transformer decoding. Using multiple model families (GPT-2, BLOOM, and OPT), we observe latency spikes of up to 21x within specific decoding-budget intervals, followed by recovery at neighboring configurations. Controlled experiments show that these anomalies originate primarily during the decode phase rather than prefill, are not explained by memory pressure alone, and remain absent on CPU and NVIDIA CUDA backends under identical conditions. We further show that key-value (KV) cache interacts strongly with these pathological execution regimes: KV caching remains beneficial overall, but its practical speedup collapses sharply within anomalous configurations, while cache-disabled decoding still exhibits residual non-monotonic behavior. These findings suggest that autoregressive decoding on MPS enters discrete execution regimes that are not captured by coarse-grained benchmarking, highlighting the importance of hardware-aware evaluation for long-context inference.