[AMBER] Uranyl Simulation Discrepancy Between CPU and GPU

From: KLEIN Jordan via AMBER <amber.ambermd.org>
Date: Tue, 4 Feb 2025 12:42:03 +0000

Dear Amber experts,

I’m trying to simulate uranyl molecules. I have a force field, and I obtain the correct uranium–oxygen distance in UO₂ (1.78 Å) when running the simulation on a CPU with pmemd. However, using the same force field on a GPU with pmemd.cuda, this distance is reduced to 1.72 Å.

In both cases, I am using AMBER 20 and a 12-6-4 LJ potential. All other distances in my simulation remain the same.

Why this distance is not the same ?


Thanks in advance.

Best regards,
Jordan

_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Tue Feb 04 2025 - 05:00:02 PST
Custom Search