Have anyone used Acellera to do a MD

rpestana94

Recently I discovered Acellera (https://www.acellera.com/), a platform to do molecular dynamic simulation, based in python, apparently, it performs really fast even for big systems, so I want to know if someone here has used it before and tell me their opinion, reliability of results, it actually that fast.

Thanks in advance.

sobereva

I don't think it is necessary to purchase this program.
The GPU accelerated GROMACS code is already extremely fast, and it has been extensively employed by MD community. Although there is no comparison, I don't believe ACEMD is significantly faster than GROMACS.

rpestana94

sobereva 发表于 2021-5-18 13:35
I don't think it is necessary to purchase this program.
The GPU accelerated GROMACS code is already ...

Thanks professor Sob, they actually provide a free license, the paid license give you the capability on running with mutiples gpu and a few extra functionality as far I know.
I don't use gromacs yet but I have undertand as you say gromacs is really fast.

How much faster is gromacs compare to amber, a rought stimation, I usullay work with amber and for big systems (200k atoms) is not a good idea to use it.

sobereva

rpestana94 发表于 2021-5-19 04:22
Thanks professor Sob, they actually provide a free license, the paid license give you the capabili ...

For purely CPU running, gromacs is significantly faster than PMEMD module of Amber
For GPU acceleration case, the difference is not so huge
200k atoms is never large (at least for GROMACS) if you have a high-end GPU card.

Compared to ACEMD, the advantage of GROMACS is not only completely free, but also GROMACS is much more mature and provides much richer features (and there are many third-part tools compatible for GROMACS).

rpestana94

sobereva 发表于 2021-5-20 05:04
For purely CPU running, gromacs is significantly faster than PMEMD module of Amber
For GPU accele ...

Thanks professor Sob