OxyGen: Unified KV Cache Management for Vision-Language-Action Models under Multi-Task Parallelism

arXiv:2603.14371v1 Announce Type: cross
Abstract: Embodied AI agents increasingly require parallel execution of multiple tasks, such as manipulation, conversation, and memory construction, from shared observations under distinct time constraints. Recent Mixture-of-Transformers (MoT) Vision-Language-Action Models (VLAs) architecturally support such heterogeneous outputs, yet existing inference systems fail to achieve efficient multi-task parallelism for on-device deployment due to redundant computation and resource contention. We identify isolated KV cache management as the root cause. To address this, we propose unified KV cache management, an inference paradigm that treats KV cache as a first-class shared resource across tasks and over time. This abstraction enables two key optimizations: cross-task KV sharing eliminates redundant prefill of shared observations, while cross-frame continuous batching decouples variable-length language decoding from fixed-rate action generation across control cycles. We implement this paradigm for $pi_0.5$, the most popular MoT VLA, and evaluate under representative robotic configurations. OxyGen achieves up to 3.7$times$ speedup over isolated execution, delivering over 200 tokens/s language throughput and 70 Hz action frequency simultaneously without action quality degradation.