cuda::abs_diff

Defined in the <cuda/cmath> header.

namespace cuda {

template <class T>
[[nodiscard]] __host__ __device__ constexpr
cuda::std::make_unsigned_t<T> abs_diff(T lhs, T rhs) noexcept;

} // namespace cuda

The function cuda::abs_diff computes absolute difference of two integers and returns the result as an unsigned value.

Parameters

• lhs: The left-hand side input.

• rhs: The right-hand side input.

Return value

Returns a cuda::std::make_unsigned_t<T> value of lhs and rhs absolute difference.

Constraints

• T must be an integer type.

Performance considerations

• On device, for types whose width is smaller than 64-bit results in:

  • VABSDIFF on SM75, SM80, SM86, SM87, SM89, SM90, SM100, SM103, and SM110.

• Other target/type combinations result in max(lhs, rhs) - min(lhs, rhs) equivalent.

Example

#include <cuda/cmath>
#include <cuda/std/cassert>
#include <cuda/std/cstdint>
#include <cuda/std/limits>

__global__ void kernel()
{
    auto lhs = cuda::std::numeric_limits<cuda::std::int32_t>::min();
    auto rhs = cuda::std::numeric_limits<cuda::std::int32_t>::max();

    assert(cuda::abs_diff(lhs, rhs) == cuda::std::numeric_limits<cuda::std::uint32_t>::max());
}

int main()
{
    kernel<<<1, 1>>>();
    cudaDeviceSynchronize();
}

See it on Godbolt 🔗