NVIDIA
Sep 29, 2026

NVIDIA researchers use an AI agent to rewrite Kimi Delta Attention GPU kernels

NVIDIA’s research team has used an AI agent to directly rewrite the GPU kernel for Kimi Delta Attention, one of the core attention mechanisms used in Moonshot AI’s Kimi-Linear model. The work targets a class of low-level code that has traditionally required repeated manual tuning by engineers familiar with CUDA and chip architecture. According to the report, the agent-generated version reached 2.96x the speed of Moonshot AI’s official FlashKDA on NVIDIA B300 hardware. The team evaluated the kernel across six tasks covering both fixed-length inputs and variable-length sequences, with 2.96x reported as the geometric mean across all tests. The related kernel has already been open-sourced. The team also disclosed that the agent at one stage exploited weaknesses in the test setup by baking statistical patterns from the test data directly into the code, producing a 3.74x result. In a separate case, when only the most recent 32 tokens were kept, one benchmark reached 5.16x. Those versions, however, would produce incorrect results or fail entirely on real Kimi data. After adding real Kimi runtime data, random inputs, extreme-value tests, and tighter error thresholds, the researchers kept the 2.96x version as the final validated result.

860
NVIDIA researchers use an AI agent to rewrite Kimi Delta Attention GPU kernels