# Kimi K3 写 H100 CUDA 内核速度比优化 PyTorch 快 14.82 倍

- 来源：Rohan Paul (@rohanpaul_ai)
- 发布时间：2026-07-19 11:13
- AIHOT 分数：34
- AIHOT 链接：https://aihot.virxact.com/items/cmrr8q87k0354bi18fho0kwqg
- 原文链接：https://x.com/rohanpaul_ai/status/2078679499320013136

## AI 摘要

Kimi K3 在 KernelBench 测试中，生成 H100 CUDA 内核的速度比优化后的 PyTorch 快 14.82 倍，性能几乎追平 Claude Opus 4.8。KernelBench 要求 AI 智能体基于可读的 PyTorch 实现，生成既正确又更快的自定义 GPU 实现，代码必须通过编译、数值正确性验证并在实际硬件上超越基准。

## 正文

Kimi K3 now writes a H100 CUDA kernel 14.82x faster than optimized PyTorch, nearly matching Claude Opus 4.8.

KernelBench tests hardware work, not just whether generated code successfully compiles.

KernelBench gives an AI agent a readable PyTorch implementation and asks it to produce a custom GPU implementation that is both correct and faster.

The code must compile, return the right numerical results across test inputs and beat a reference implementation on actual hardware. A kernel that is fast but wrong fails. A kernel that is correct but slow is not particularly useful.

### 引用推文

> Xinyu Yang：Now you can ask K3 to write the kernels From https://kernelbench.com