[gmx-users] Hyper-threading on multi-CPU xeon worstation

[Gleb Novikov]

Dear Gromacs users!
I wonder to ask whether is possible to expect an increase in Gromacs performance with HT technology on my Xeon cpu (72 cpus without HT and 144 with HT)? How would be better to calibrate mdp options  (e.g integrator, adjust electrostatics cut-offs) to obtain best performance from my workstation with enabled HT? Does the HT give better performance in case of simulation of big systems (e.g MARTINI  with  > 200 of CG particles)?

Thanks in advance!
Gleb