I am not too familiar with Theano, but from what I can tell it is more focused towards Deep Learning. So I will refrain from comparing and will give you a short list about ArrayFire.
- Supports multiple backends, so you can run on NVIDIA GPUs, AMD GPUs, Intel Xeon Phis, and all CPUs using the same API.
- ArrayFire currently has statistics, image processing, signal processing and Linear algebra functions. We are planning to add Machine Learning and Computer Vision functions / algorithms in the near future.
- ArrayFire is a native (C/C++) library. It can be used from other languages fairly easily.
- The main goal is to make parallel programming in general (GPU programming in particular) easier and portable.
- Supports multiple backends, so you can run on NVIDIA GPUs, AMD GPUs, Intel Xeon Phis, and all CPUs using the same API.
- ArrayFire currently has statistics, image processing, signal processing and Linear algebra functions. We are planning to add Machine Learning and Computer Vision functions / algorithms in the near future.
- ArrayFire is a native (C/C++) library. It can be used from other languages fairly easily.
- The main goal is to make parallel programming in general (GPU programming in particular) easier and portable.