[x265] Some questions about x265

[GOU Lingfeng]

Dear Developers,

I'm looking for a method to compress volume data (stacks of images, one
data set terabytes in size when uncompressed).
After some research, I found that HEVC is most efficient.
I tried the reference implementation (HM, 
https://hevc.hhi.fraunhofer.de/) and x265.
HM is so slow that we have not enough computational resources to
practically compress a whole data set.
On the other hand, x265 lacks 16-bit support, which is required for
some data sets (16-bit gray-scale).

1. Is it likely to make HM run several times faster, by tuning encoder
parameters?
2. Or by replacing time consuming codes in HM with faster
implementation? (What makes x265 so much faster?)
3. Is it possible to extend x265 with 16-bit gray-scale support? How
complicated will it be?

Thank you so much!


Sincerely,
GOU Lingfeng