Google Cloud GPUs

What do you like best?

- Seamless integration of GPUs with any VM - easily attach one or more GPUs to instances to accelerate specific workloads and offload work from CPUs.

- Choice of GPU(s) - from a single GPU to multiple virtual GPU cluster.

- Tensorflow and Keras work seamlessly with the GPU powered VMs as they have preinstalled drivers; a great selection of images to bootstrap the VM

- Ample documentation and samples to work with GPUs in their Vs

What do you dislike?

- The pricing is expensive compared to competitors (AWS and Azure)

- Customer support is not great - need to purchase a Silver, Gold, or Platinum package for better support.

What problems are you solving with the product? What benefits have you realized?

- Machine learning - Image and NLP

- Training models

- Image processing