Is 'read_imagef' with sampler with 'CLK_FILTER_LINEAR' option hardware accelerated?

Hi.
My platforms are Intel HD Graphics 5500 and Nvidia Cuda from Geforce 645.
Using those devices on OpenCL, does ‘read_imagef’ with sampler with ‘CLK_FILTER_LINEAR’ leverage hardware accelerated bilinear interpolation?

My understanding is “yes”.