Re: [AMBER] RTX 5000 cards and AMBER 16 compatibility

From: Hashemi, Mohtadin <mohtadin.hashemi.unmc.edu>
Date: Sat, 3 Aug 2019 03:38:12 +0000



On Aug 2, 2019, at 16:17, Laura Furge <Laura.Furge.kzoo.edu<mailto:Laura.Furge.kzoo.edu>> wrote:

Non-UNMC email

I am looking to buy a new machine for running simulations and am doing some comparisons right now. I want to keep AMBER 16 for some technical reasons. My question - does anyone know if AMBER 16 is compatible with RTX 5000 cards from Dell?

If you are planning to use CUDA ver. 10 then it’s not officially supported for Amber16 (http://ambermd.org/gpus16/index.htm). But you should be able to use the card with CUDA 9.x, you may not get the same utilization of the GPU though (https://docs.nvidia.com/cuda/turing-compatibility-guide/index.html).


1.3. Compatibility between Volta and Turing<https://docs.nvidia.com/cuda/turing-compatibility-guide/index.html#turing-volta-compatibility>

The Turing architecture is based on Volta's Instruction Set Architecture ISA 7.0, extending it with new instructions. As a consequence, any binary that runs on Volta will be able to run on Turing (forward compatibility), but a Turing binary will not be able to run on Volta. Please note that Volta kernels using more than 64KB of shared memory (via the explicit opt-in, see CUDA C Programming Guide) will not be able to launch on Turing, as they would exceed Turing's shared memory capacity.

Most applications compiled for Volta should run efficiently on Turing, except if the application uses heavily the Tensor Cores, or if recompiling would allow use of new Turing-specific instructions. Volta's Tensor Core instructions can only reach half of the peak performance on Turing. Recompiling explicitly for Turing is thus recommended.”


If you want to use CUDA ver. 10 then you need to modify the config file. We have done so for a test and did not see any problems and or errors running our test suite. I am not recommending this nor do I take any responsibility for any and all problems and or errors should anyone do this.


Thank you for any insights,
Laura



--
Laura Lowe Furge, Ph.D.
Associate Provost
Dorothy H. Heyl Professor of Chemistry
Kalamazoo College
1200 Academy Street
Kalamazoo, MI 49006
(she/her/hers)
email: Laura.Furge.kzoo.edu<mailto:Laura.Furge.kzoo.edu>
Phone: 269.337.7162
FAX: 269.337.7508
website: https://urldefense.proofpoint.com/v2/url?u=http-3A__reason.kzoo.edu_chem_faculty_laura_&d=DwICAg&c=ZukO2fIan9e5E9v43wuy1w&r=AK606AZl4zLFP6DQFHRjBg0DSGYp515YmJqZ3zuhSTk&m=LHnY21ae0LI5J_TYOhd5VOgFLlPKls4bk-mlJDZMMus&s=-1U9GOEIUpQ6xof7xEacGn_Clw6v2I9UC0N8JAfvoBI&e=

My office is in a location that may be physically inaccessible to chair users or people with other mobility-related disabilities. If you need to meet in an alternate location, please let me know.
_______________________________________________
AMBER mailing list
AMBER.ambermd.org<mailto:AMBER.ambermd.org>
https://urldefense.proofpoint.com/v2/url?u=http-3A__lists.ambermd.org_mailman_listinfo_amber&d=DwICAg&c=ZukO2fIan9e5E9v43wuy1w&r=AK606AZl4zLFP6DQFHRjBg0DSGYp515YmJqZ3zuhSTk&m=LHnY21ae0LI5J_TYOhd5VOgFLlPKls4bk-mlJDZMMus&s=h5IUexa_CUgdoPalO3H-r2A6Dox_Mtheq_h-Ldk7XqM&e=

The information in this e-mail may be privileged and confidential, intended only for the use of the addressee(s) above. Any unauthorized use or disclosure of this information is prohibited. If you have received this e-mail by mistake, please delete it and immediately contact the sender.
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Fri Aug 02 2019 - 21:00:02 PDT
Custom Search