[gmx-users] GROMACS vs. AMBER speed comparison?

[Mark Abraham]

Tsjerk Wassenaar wrote:
> Hi Seungpyo Hong,
> 
>> Hi, gmx-users,
>>
>> Recently I have try to use AMBER 9 for the simulation of protein with about
>> 300 residues.
>> Surprisingly, I found that AMBER is far slower than GROMACS; GROMACS is
>> about twice faster.
>> I can not figure out the reason.
>> Why is GROMACS faster than AMBER?
>> Could anybody know the reason?
> 
> The reason is that the developers wanted it to be faster and put time
> in thinking about the routines. The actual cause of the differences in
> speed are well described in the special issue of JCC, december 2005 on
> MD packages, and on the web site, as suggested by Yang Ye.

JCTC December 2005, IIRC :-)

Mark