@ cuda.jit和@jit之间的区别(target ='gpu') [英] Difference between @cuda.jit and @jit(target='gpu')

问题描述

我对使用Continuum的Accelerate和numba软件包中的Python CUDA库有疑问.将装饰器 @jit 与 target = gpu 一起使用是否与 @ cuda.jit 相同?

I have a question on working with Python CUDA libraries from Continuum's Accelerate and numba packages. Is using the decorator @jit with target = gpu the same as @cuda.jit?

推荐答案

不，尽管最终编译成PTX的编译路径是相同的，但它们并不相同. @jit 装饰器是常规的编译器路径，可以选择将其引导到CUDA设备上. @ cuda.jit 装饰器实际上是Continuum Analytics开发的低级Python CUDA内核方言.因此，您将获得对 @ cuda.jit 中的CUDA内置变量(如 threadIdx )和内存空间说明符(如 __ shared __ )的支持.

No, they are not the same, although the eventual compilation path into PTX into assembler is. The @jit decorator is the general compiler path, which can be optionally steered onto a CUDA device. The @cuda.jit decorator is effectively the low level Python CUDA kernel dialect which Continuum Analytics have developed. So you get support for CUDA built-in variables like threadIdx and memory space specifiers like __shared__ in @cuda.jit.

如果要用Python编写CUDA内核并进行编译和运行，请使用 @ cuda.jit .否则，如果您想加速现有的Python，请对CUDA目标使用 @jit .

If you want to write a CUDA kernel in Python and compile and run it, use @cuda.jit. Otherwise, if you want to accelerate an existing piece of Python use @jit with a CUDA target.

这篇关于@ cuda.jit和@jit之间的区别(target ='gpu')的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！