Page 17 of 17

Re: WRF EMS Benchmark results and hardware

Posted: Tue Jun 23, 2015 12:02 pm
by Larf
Thanks for your estimations, meteoadriatic.
I agree, it shouldn´t take longer, by comparison though - Europe should have something like 4000 nodes - and (in my case) 10 years take 86 days of calculation, results in a 1000 year run!!
. Presumably we`ll have a -not yet defined - work Station running in September. I'ma post the Benchmark results.
BTW, that is what I figured, too, windyweek: having memory channels per core seems to be a very quickening aspect.