3.11. GPU Reduction

Writing a reduction algorithm for CUDA GPU can be tricky. Numba provides a @reduce decorator for converting simple binary operation into a reduction kernel.

3.11.1. @reduce


import numpy
from numba import cuda

def sum_reduce(a, b):
    return a + b

A = (numpy.arange(1234, dtype=numpy.float64)) + 1
expect = A.sum()      # numpy sum reduction
got = sum_reduce(A)   # cuda sum reduction
assert expect == got

User can also use a lambda function:

sum_reduce = cuda.reduce(lambda a, b: a + b)

3.11.2. class Reduce

The reduce decorator creates an instance of the Reduce class. (Currently, reduce is an alias to Reduce, but this behavior is not guaranteed.)

class numba.cuda.Reduce(binop)
__call__(arr, res=None, size=None, init=0, stream=0)

Performs a full reduction.

  • arr – A host or device array. If a device array is given, the reduction is performed inplace and the values in the array are overwritten. If a host array is given, it is copied to the device automatically.
  • size – Optional integer specifying the number of elements in arr to reduce. If this parameter is not specified, the entire array is reduced.
  • res – Optional device array into which to write the reduction result to. The result is written into the first element of this array. If this parameter is specified, then no communication of the reduction output takes place from the device to the host.
  • init – Optional initial value for the reduction, the type of which must match arr.dtype.
  • stream – Optional CUDA stream in which to perform the reduction. If no stream is specified, the default stream of 0 is used.

If res is specified, None is returned. Otherwise, the result of the reduction is returned.


Create a reduction object that reduces values using a given binary function. The binary function is compiled once and cached inside this object. Keeping this object alive will prevent re-compilation.

Parameters:binop – A function to be compiled as a CUDA device function that will be used as the binary operation for reduction on a CUDA device. Internally, it is compiled using cuda.jit(signature, device=True).