PIXELBANKv8.2.1
Menu
Sign In

CUDA Vector Addition Kernel

EasyCUDA Execution Model

Problem Statement

Write a CUDA kernel that adds two 1D float arrays element-wise: c = a + b.

Background

Every CUDA thread computes its own global index. With Numba you get it from cuda.grid(1) (equivalent to blockIdx.x * blockDim.x + threadIdx.x). One thread handles one element, so the launch needs ceil(n / threads) blocks.

Your Task

Implement add_kernel and a run(n=1024) that launches it over a 1D grid and returns whether the result matches a + b.

How it is tested

Your solution must define a top-level function run(...) that allocates the inputs, copies them to the GPU, launches your @cuda.jit kernel, and returns a Python bool from np.allclose(gpu_result, reference). The grader prints run(...); the expected output is True.

Example:

Input:
n = 1024
Output:
True
Reasoning:
  • The input n = 1024 determines the size of the input arrays a and b, which are 1D float arrays of length n=1024n = 1024.
  • The add_kernel function is launched over a 1D grid with ceil(n / threads) blocks, where each thread handles one element of the arrays, performing element-wise addition: ci=ai+bic_i = a_i + b_i.
  • The result of the kernel launch is stored in the gpu_result array, which is then compared to the reference result computed using NumPy: reference=a+breference = a + b.
  • The final output is True if the two results match within a tolerance, as determined by np.allclose(gpu_result, reference).

Constraints:

  • Use @cuda.jit and i = cuda.grid(1)
  • Guard with if i < a.size before writing
  • blocks = (n + threads - 1) // threads
Editor

Test Results

0/0
Run code to see test results.