From: Norman Geist (norman.geist_at_uni-greifswald.de)
Date: Mon Mar 21 2016 - 03:31:30 CDT
Ok in that case: The STMV and APOA1 benchmarks are widely used as reference for large and medium sized systems. Scaling one and the same system isn’t that straight forward, or reasonable. If you want an easy to scale system, you could build systems of just water for different box sizes, this can be done using VMD and the Solvate plugin for instance. In case you want to benchmark on GPUs, you may want to test the impact of “twoaway[xyz] yes” consecutive in the namd job script aswell as +ignoresharing on the commandline.
Best of luck
Norman Geist
Von: Eric Germaneau [mailto:germaneau_at_sjtu.edu.cn]
Gesendet: Montag, 21. März 2016 09:15
An: Norman Geist <norman.geist_at_uni-greifswald.de>
Cc: namd-l_at_ks.uiuc.edu
Betreff: Re: AW: namd-l: Testing NAMD scalability
Thanks,
I didn't say I don't have any MD experience, I'm just not a namd user.
Sent using CloudMagic Email <https://cloudmagic.com/k/d/mailapp?ct=pa&cv=8.3.8&pv=4.4.4&source=email_footer_2>
On Mon, Mar 21, 2016 at 3:46 PM, Norman Geist <norman.geist_at_uni-greifswald.de <mailto:norman.geist_at_uni-greifswald.de> > wrote:
Similar question have been asked in the past. Maybe search the mailing list archive. Away from that, the NAMD folks do offer benchmark systems of various size for download, which are a good point to start.
BTW. as I pointed out in earlier posts, I highly doubt that anybody without MD experience can carry out reliable benchmarks.
Norman Geist
Von: owner-namd-l_at_ks.uiuc.edu <mailto:owner-namd-l_at_ks.uiuc.edu> [mailto:owner-namd-l_at_ks.uiuc.edu] Im Auftrag von Éric Germaneau
Gesendet: Samstag, 19. März 2016 02:32
An: namd-l_at_ks.uiuc.edu <mailto:namd-l_at_ks.uiuc.edu>
Betreff: namd-l: Testing NAMD scalability
Dear all,
I'm not a NAMD user but I nned to test its scalability on our facility.
Can someone tell me how to generate different size of the same system?
Can I take any of those case studies <https://www-s.ks.uiuc.edu/Training/CaseStudies/> and scale them?
Thank you.
-- Éric Germaneau (艾海克), Specialist Center for High Performance Computing Shanghai Jiao Tong University Room 205 Network Center, 800 Dongchuan Road, Shanghai 200240 China Email:germaneau_at_sjtu.edu.cn Mobi:+86-136-4161-6480 http://hpc.sjtu.edu.cn
This archive was generated by hypermail 2.1.6 : Sun Dec 31 2017 - 23:20:14 CST