Supercomputers for Weather Forecasting Have Come a Long Way

Re:Supercomputers for Weather Forecasting Have Come a Long Way by gbrown_ on Wednesday March 01, 2017 @10:51AM (#13764638)

    Currently the most powerful weather forecasting computer is the UK Met Office
    machine, a Cray XC40 machine that delivers 6.8 petaflops and sits at number 11
    on the TOP500.

But this does not run the production forecast, that is done on one of two smaller machines which are capable just over 3 petaflops.

Source: I work for Cray at the UK Met Office.

[ Parent | on HN ]

Re:Supercomputers for Weather Forecasting Have Come a Long Way by gingerbread-man on Wednesday March 01, 2017 @10:53AM (#13764658)

So the XC40 is used for research purposes? I'd be very interested to know more about your workflow.

[ Parent | on HN ]

Re:Supercomputers for Weather Forecasting Have Come a Long Way by gbrown_ on Wednesday March 01, 2017 @10:55AM (#13764677)

That XC40 system is for collaboration work but it also serves as a test bed of sorts as it's so much lager users can try scaling up models on it.

[ Parent | on HN ]

Re:Supercomputers for Weather Forecasting Have Come a Long Way by alex_hitchins on Wednesday March 01, 2017 @10:57AM (#13764708)

What is the 'bigger' one used for then?

Edit : Answered while I was posting this

[ Parent | on HN ]

Re:Supercomputers for Weather Forecasting Have Come a Long Way by crosbyar on Wednesday March 01, 2017 @11:49AM (#13765384)

This is a very good point. Large clusters are typically shared resources, this is a big problem for the US forecasts since we have to balance the time/cpu and timeliness of results.

[ Parent | on HN ]

Re:Supercomputers for Weather Forecasting Have Come a Long Way by gbrown_ on Wednesday March 01, 2017 @11:58AM (#13765520)

Well production runs on one of two identical machines. This kind of redundancy is not uncommon for weather sites. One of these will run the production version of the forecast and the other will be running version n+1. We flip between the two semi-regularly to ensure if one machine were to have a major issue production would be unaffected if forced to move to the other machine. Smaller weather sites may have a single machine partitioned into two.

[ Parent | on HN ]

Re:Supercomputers for Weather Forecasting Have Come a Long Way by NLips on Wednesday March 01, 2017 @10:54AM (#13764667)

Please can you tell us what forecasts it is running then?

[ Parent | on HN ]

Re:Supercomputers for Weather Forecasting Have Come a Long Way by onion2k on Wednesday March 01, 2017 @10:54AM (#13764666)

What does it do then?

[ Parent | on HN ]

Re:Supercomputers for Weather Forecasting Have Come a Long Way by popobobo on Wednesday March 01, 2017 @11:34AM (#13765155)

Is it running fortran? Which version?

[ Parent | on HN ]

Re:Supercomputers for Weather Forecasting Have Come a Long Way by batbomb on Wednesday March 01, 2017 @11:48AM (#13765375)

It runs Cray Linux (based on SUSE) and assuredly somebody is running Fortran on it in some form.

Many people think these large machines are intended for a small handful of users that run extremely large jobs. That's typically not the case; they typically have tens of users per day up to hundreds per week, depending on institution. With diverse users comes diverse software and the challenges based on that.

[ Parent | on HN ]

Re:Supercomputers for Weather Forecasting Have Come a Long Way by gbrown_ on Wednesday March 01, 2017 @11:42AM (#13765272)

Not sure exactly what you're after here. The systems have assorted versions of Fortran compilers from both Cray and Intel. With regards to versions of the language standard used/ required for the models I'm not sure.

[ Parent | on HN ]