Hello AmberTools,
We've currently got some time on the Bristol (UK) super cluster, which has
specifications here —
https://docs.isambard.ac.uk/specs/#system-specifications-isambard-ai-phase-2.
Essentially, it's an aarch64 architecture composed of NVIDIA GH200 Grace
Hopper Superchips, the C++ arch flag is `-mcpu=neoverse-v2` if that helps.
I'm attempting to compile the library for that target, but I wanted to
check if there have been any previous efforts to do this. I see an old
AmberTools v20 on Spack which sadly doesn't install on this cluster —
https://packages.spack.io/package.html?name=amber — as I don't think it's
built for aarch64. So I figure that contributing a new version is probably
the way to go. I'm not familiar with this library, but somewhat familiar
with CMake and C++ I'll be looking to compile AmberTools with support for
MPI and CUDA.
Who would be knowledgable on this subject? We're happy to look at
sponsoring this work if that's preferable.
Best regards,
Emlyn Clay
Director, Software Engineering
https://www.xyme.ai/
--
_The content of this email is intended for the person or entity to which it
is addressed only. This email may contain confidential information. If you
are not the person to whom this message is addressed, be aware that any
use, reproduction, or distribution of this message is strictly prohibited.
If you received this in error, please contact the sender and immediately
delete this email and any attachments.
_
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed Nov 12 2025 - 10:30:02 PST
