Re: Do firsttimestep, numsteps and run have maximum values?

From: Brian Radak (brian.radak_at_gmail.com)
Date: Wed Sep 19 2018 - 09:32:29 CDT

I'm 90% sure these are all 32 bit integers in the code. It's also possible
that there is not much overflow checking until a logical error is
encountered.

What do you specifically have to gain from running so many steps at once? I
can't imagine such a procedure is robust on most hardware.

Firsttimestep is also just a bookkeeping device and doesn't impact output
beyond simple mod operations for flagging output.

On Tue, Sep 18, 2018, 9:44 PM yjcoshc <yjcoshc_at_gmail.com> wrote:

> Hello everyone,
>
> I am trying to restart a NAMD simulation which has run for 2133000000
> steps. If I set:
>
> firsttimestep 2133000000
> run 1000000000
>
> and then NAMD would run 0 step and exit. I have looked at the NAMD
> source code and found the type of N and firsttimestep is int, not long
> long or int64_t. Does that mean the maximum firsttimestep and total
> steps is the maximum int type integer (2147483647 in gcc)?
>
> Thanks,
>
> Haochuan Chen
>
>

This archive was generated by hypermail 2.1.6 : Wed Dec 04 2019 - 23:20:12 CST