Thanks for testing it out. There are probably more tuning opportunities for fortran (e.g. larger solution search space, more aggressive pruning, and more advanced loop invariants and register pressure estimation), which I hope someone can continue working on (or me if I find more time).
David On Thu, Jul 29, 2010 at 11:33 AM, Toon Moene <t...@moene.org> wrote: > It buys the HIRLAM code about 100 seconds out of 8 thousands: > > New: > > $ grep 'FORECAST TOOK' HL_Cycle_2010072912.html > FORECAST TOOK 775.5685 SECONDS > FORECAST TOOK 779.8127 SECONDS > FORECAST TOOK 29.8419 SECONDS > FORECAST TOOK 7929.5913 SECONDS > > Compared to (old): > > $ grep 'FORECAST TOOK' HL_Cycle_2010072812.html > FORECAST TOOK 785.2891 SECONDS > FORECAST TOOK 786.7651 SECONDS > FORECAST TOOK 31.5340 SECONDS > FORECAST TOOK 8027.6938 SECONDS > > -- > Toon Moene - e-mail: t...@moene.org - phone: +31 346 214290 > Saturnushof 14, 3738 XG Maartensdijk, The Netherlands > At home: http://moene.org/~toon/; weather: http://moene.org/~hirlam/ > Progress of GNU Fortran: http://gcc.gnu.org/gcc-4.5/changes.html#Fortran >
