Hello Ross,
Excellent, thank you very much.
On a side question, the current machine I am running on (not my research station) has older hardware to support the amber18. What is considered a "good enough" box size and cut-off for a protein-DNA MD simulation. As I have run some apo- MD sims with a large box (16 Angstrom) and cut-off (14 Angstrom), but the protein-DNA MD sims will take considerably longer, so I was wondering if you could reduce the box size or cut-off or would this disallow comparisons between the apo- system and the protein-DNA system?
Regards,
Dan.
From: Ross Walker <rosscwalker.gmail.com> On Behalf Of Ross Walker
Sent: Thursday, March 23, 2023 2:38 PM
To: Daniel Hall <D.Hall1.bradford.ac.uk>
Cc: AMBER Mailing List <amber.ambermd.org>
Subject: Re: [AMBER] AMBER18 pmemd.cuda installation on NVIDIA 3080
Caution External Email: Do not click any links or open any attachments unless you trust the sender and know that the content is safe.
Hi Daniel,
http://ambermd.org/GetAmber.php#amber
At present it's $500 but it will become free for academic users from April 15th 2023 onwards. As such you can either wait until April 15th or you could contact dacase.ambermd.gmail.com<mailto:dacase.ambermd.gmail.com> and see if you can get a fee waiver in the interim.
All the best
Ross
On Mar 23, 2023, at 10:26, Daniel Hall <D.Hall1.bradford.ac.uk<mailto:D.Hall1.bradford.ac.uk>> wrote:
Hello Ross,
Thank you so much for your prompt response. How much is the upgrade fee from AMBER18 to AMBER22 for a UK university institution?
Regards,
Dan.
-----Original Message-----
From: Ross Walker <rosscwalker.gmail.com<mailto:rosscwalker.gmail.com>> On Behalf Of Ross Walker
Sent: Thursday, March 23, 2023 2:22 PM
To: Daniel Hall <D.Hall1.bradford.ac.uk<mailto:D.Hall1.bradford.ac.uk>>; AMBER Mailing List <amber.ambermd.org<mailto:amber.ambermd.org>>
Subject: Re: [AMBER] AMBER18 pmemd.cuda installation on NVIDIA 3080
Caution External Email: Do not click any links or open any attachments unless you trust the sender and know that the content is safe.
Hi Daniel,
Your premise is correct - the RTX30XX series of GPUs are not compatible with Amber 18 since they were released several years after Amber 18 was. I suggest upgrading to Amber 22 which will support it. I would also recommend using CUDA 11.8 (since CUDA 10.2 is too old for the RTX3080). Make sure you have the latest NVIDIA driver as well and that nvidia-smi -a correctly identifies the 3080.
All the best
Ross
On Mar 23, 2023, at 10:16, Daniel Hall via AMBER <amber.ambermd.org<mailto:amber.ambermd.org>> wrote:
Hello AMBER users,
I am trying to utilize the NVIDIA 3080 GPU I have installed in my workstation for MD simulations. I have managed to run AMBER Serial and Parallel with use of OPENMPI version 2.1.6 with no problems but I am wanting to run pmemd.cuda with my GPU. I am using the compiler gcc 4.8.5, CUDA 10.2 and running the command
./configure -cuda -noX11 gnu
When I try run
make test
This outputs
0 file comparisons passed
15 file comparisons failed
369 tests experienced errors
I think I have set up the paths correctly
CUDA_HOME=/usr/local/cuda-10.2
AMBERHOME=/home/dhall1/amber/amber18
export LD_LIBRARY_PATH=/usr/local/cuda-10.2/lib64:$LD_LIBRARY_PATH
The current error message I get when I try and run the pmemd.cuda command is
cudaMemcpyToSymbol: SetSim copy to cSim failed invalid device symbol
My question is firstly, is the NVIDIA 3080 GPU compatible with AMBER18, secondly what I am doing wrong if AMBER18 is compatible with this GPU?
Thank you in advance for any response.
Regards,
Dan.
_______________________________________________
AMBER mailing list
AMBER.ambermd.org<mailto:AMBER.ambermd.org>
https://urldefense.com/v3/__http://lists.ambermd.org/mailman/listinfo/amber__;!!Mih3wA!HorEem50gjXrLExLiP4yBKk-4udCzBWDJOwmryZ1bIBloWXUjvVJgkINMyuwOmKSyL6t52ehWW0$<
https://urldefense.com/v3/__http:/lists.ambermd.org/mailman/listinfo/amber__;!!Mih3wA!HorEem50gjXrLExLiP4yBKk-4udCzBWDJOwmryZ1bIBloWXUjvVJgkINMyuwOmKSyL6t52ehWW0$>
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Thu Mar 23 2023 - 08:00:05 PDT
