AI 导读
爆米花粒是玉米的一个特殊品种的种子，加热后能爆开并翻出内部。而 kernel 也是几百行 GPU 代码，负责执行 LLM 中的数学运算和注意力操作。GPU 只有在 kernel 持续喂饱 tensor core 时才能发挥其标称 FLOPS。草率的显存访问模式会导致芯片空转。在 OpenAI 的服务规模下，个位数百分比的 kernel 提升就意味着数亿美元的计算成本。(1/2)🧵
正文
A (popcorn) kernel is a specialized seed from a variety of corn that can burst open and turn inside out when heated. A kernel is also a few hundred lines of GPU code that executes the math and attention op in an LLM. A GPU only delivers its paper FLOPS if kernel keeps the tensor cords fed. Sloppy memory access patterns results in silicon that sits idle. At OpenAI's serving scale, single digit % kernel gains are hundreds of millions in compute. (1/2)🧵
来源：@SemiAnalysis_ · x.com