Added cudnn nhwc support for:
1. batch norm
2. convolution
3. convolution_transpose

suggest_memory_format has ambiguous meaning for two cases:
1. tensor with NCHW where C = 1.
   we could use stride of C as a hint to tell the intended memory format.
2. tensor with NCHW where H == W == 1.
   there's no way to identify the intended memory format from strides.

Currently we fallback to NCHW whenever we see contiguous tensor. Hence avoiding
ambiguity for some of the special cases.

The old implementation assumed `is_channels_last_contiguous_` to be mutually
exclusive to `is_contiguous_`, which is not true.
Properly set the flag by checking strides.

…e_empty_supports_memory_format

Initial kernel support added for optimized NHWC tensor.

TODO: currently backwards kernel spits out tensor with NHWC stride.
Unfortunately autograd restores grad to contiguous (in either copy or add). This
makes real perf tuning annoying to do. (since I cannot easily measure end-to-end
time in my python script)

My current kernel is blazing fast comparing to the original NCHW kernel in fp16,
since I avoided atomicAdd. I'll finish perf tuning after we merged some future
PR expanding NHWC support in the core.

This reverts commit c7ece81.

Previous kernel does not stride on Channel dimension, and the kernel uses shared
memory to store temporary result (to break data dependency -> code paralellism)

Resulted in requesting more resources than what's available.

Fixing:
added striding on C to reduce shmem usage per CTA.

Updated cudnn API for batchnorm. Enabling the Extended API which provides
semi-persistent batchnorm kernel that has better performance on NHWC layout.

TODO: I made adjustments to the API as well as BN in JIT IR. But I haven't fully
tested the JIT part yet. I should verify that in the final PR.