Seeking information on low-level TPU interaction and libtpu.so API

[notlober]

I'm looking to build an automatic differentiation library for TPUs without using high-level front-ends like TensorFlow/JAX/PyTorch-XLA, but I'm finding information about lower-level TPU usage is practically non-existent.

Specifically, I'm interested in:

1. How to interact with TPUs at a lower level than what's typically exposed in TensorFlow
2. Information about the libtpu.so library and its API
3. Any resources or documentation on implementing custom TPU operations

Are there any insights or suggestions on how to approach this, particularly regarding TPU support? Any ideas or help would be greatly appreciated.

I understand that some of this information might be proprietary, but any guidance on what is possible or available would be very helpful.

[sdasgup3]

related discussion on discord: https://discord.com/channels/999073994483433573/1004823487124357182/1268874394021793965