If the operating system is up and running (and that includes U-Boot, which is really an operating system), then you can write assembler, or embed assembler within a C program. If you intend to program via JTAG, then you are probably out of luck.
You’d be using ARMv8-a architecture (64-bit ARM Cortex-A series). Go here, look for the Cortex-A docs:
The CPU is a quad-core ARM Cortex-A57, so this is the document series you are interested in. I don’t remember which revision this is, and I suppose this could have changed over time, but the “MP” TRM docs are the multi-core processor docs you are interested in.
Anyone know the specific ARM Cortex-A57 revision of the TX2? Or if all units have the same revision?