Re: GA ver. Parallel tempering
- From: "Dov" <dov.bai@xxxxxxxxx>
- Date: 14 Sep 2006 19:01:28 -0700
Lionel B wrote:
To clarify the point I was making: since both SA and GAs are serial
algorithms when implemented on a serial machine, whether they are
"parallelisable" or not is irrelevant; the comparison *is* valid *on a
serial machine*. On a true parallel architecture, where solutions may be
evaluated simultaneously, the comparison would indeed *not* be fair.
I never said that parallel tempering needs to run on parallel machines
to be efficient. The word "parallel" in "parallel tempering" is not
related to the hardware but to the
algorithm.
Dov
--
Lionel B
.
- Follow-Ups:
- Re: GA ver. Parallel tempering
- From: Lionel B
- Re: GA ver. Parallel tempering
- References:
- GA ver. Parallel tempering
- From: Dov
- Re: GA ver. Parallel tempering
- From: Johann Dréo
- Re: GA ver. Parallel tempering
- From: Lionel B
- Re: GA ver. Parallel tempering
- From: Dov
- Re: GA ver. Parallel tempering
- From: Lionel B
- GA ver. Parallel tempering
- Prev by Date: Re: GA ver. Parallel tempering
- Next by Date: CEC 2007
- Previous by thread: Re: GA ver. Parallel tempering
- Next by thread: Re: GA ver. Parallel tempering
- Index(es):
Relevant Pages
|
|