AI 中文总结
FABRICA是结合目标知识、故障修复与正确性优化的智能体框架，在晶圆级系统的CUDA转CSL内核任务中，大幅提升成功率并实现3.47×以上加速，凸显跨架构内核生成的关键因素。
AI 中文摘要
将GPU内核在不同架构间移植需要架构重映射，而非语法替换。CUDA通过线程、块和内存访问编码分解、局部性与同步；Cerebras软件语言（CSL）则要求显式放置、分布式SRAM、fabric通信、事件驱动任务及主机/设备契约。本文提出FABRICA-Bench（含49个配对的CUDA到CSL任务），以及FABRICA——一种结合目标架构知识、执行、故障导向修复和正确性门控优化的智能体框架。在固定的28项Level 1-3核心任务对比中，FABRICA相比Claude Opus 4.8将成功率从6/28提升至26/28，其中22个成功程序匹配或优于其CSL参考实现。在49项任务的覆盖度评估中，38项任务生成正确程序；最后3项任务经3个随机种子评估，在9次运行中通过8次。对于27个带设备内部时序的生成/参考程序对，几何平均加速比在SDK模拟器上为3.75×，在WSE-3硬件上为3.47×。在可执行工作流固定的情况下，Claude Opus 4.8通过26/28项核心任务，而最佳开源权重模型仅通过2/28项；检索到的Cerebras知识单独将Level 1-3面板任务的成功率从1/15提升至7/15。这些结果表明，基础模型能力、目标架构知识、执行反馈和同目标测量是跨架构内核生成的关键因素。
英文摘要
Porting GPU kernels across architectures requires architectural remapping, not syntax substitution. CUDA encodes decomposition, locality, and synchronization through threads, blocks, and memory accesses; the Cerebras Software Language (CSL) requires explicit placement, distributed SRAM, fabric communication, event-driven tasks, and host/device contracts. We present FABRICA-Bench, 49 paired CUDA-to-CSL tasks, and FABRICA, an agentic framework combining target knowledge, execution, failure-directed repair, and correctness-gated optimization. On a fixed 28-task Level~1--3 core comparison with Claude Opus 4.8, FABRICA raises success from 6/28 to 26/28; 22 successful programs match or beat their CSL references. Across the 49-task coverage evaluation, 38 tasks produce a correct program; the final three tasks are evaluated over three seeds and pass 8/9 runs. For 27 generated/reference pairs with device-internal timing, geometric-mean speedup is 3.75$\times$ on the SDK simulator and 3.47$\times$ on WSE-3 hardware. With the executable workflow fixed, Claude Opus~4.8 passes 26/28 core tasks while the best open-weight model passes 2/28; retrieved Cerebras knowledge separately raises success from 1/15 to 7/15 on a Level~1--3 panel. These results identify base-model capability, target knowledge, execution feedback, and same-target measurement as central to cross-architecture kernel generation.
