Nightjar: Dynamic Adaptive Speculative Decoding for Large Language Models Serving
This work addresses a critical bottleneck in real-time LLM serving by optimizing resource usage under dynamic workloads, though it is incremental as it builds on existing speculative decoding techniques.
The paper tackles the performance trade-off in speculative decoding for LLM serving, where fixed methods degrade under high load, and proposes Nightjar, a dynamic adaptive framework that adjusts speculative length and disables speculation when unbeneficial, achieving an average 27.29% higher throughput and up to 20.18% lower latency compared to standard methods.
Speculative decoding (SD) accelerates LLM inference by verifying draft tokens in parallel. However, this method presents a critical trade-off: it improves throughput in low-load, memory-bound systems but degrades performance in high-load, compute-bound environments due to verification overhead. Existing speculative decoding methods use fixed lengths and cannot adapt to workload changes or decide when to stop speculation. The cost of restarting speculative inference also remains unquantified. Under high load, the benefit of speculation diminishes, while retaining the draft model reduces KV-cache capacity, limiting batch size and degrading throughput. To overcome this, we propose Nightjar, a resource-aware adaptive speculative framework. It first adjusts to the request load by dynamically selecting the optimal speculative length for different batch sizes. Crucially, Nightjar proactively disables speculative decoding when the MAB planner determines that speculation is no longer beneficial, and during the disabled phase, offloads the draft model to the CPU only under GPU memory pressure. This reclaims memory for the KV cache, thereby facilitating larger batch sizes and maximizing overall system throughput. Experiments show that Nightjar achieves average 27.29% higher throughput and up to 20.18% lower latency compared to standard speculative decoding under dynamic request arrival rates in real-time LLM serving scenarios.