AI-Assisted GPU Porting of a 250k Line Legacy Weather Simulation Code arxiv.org 56 points by Jimmc414 3 days ago
acidburnNSA 3 days ago Oh man I have to try this on all the legacy Fortran nuclear reactor simulator codes originally written in the 1960s dboreham 3 days ago I thought that's what "legacy weather simulation code" was! fooker 2 days ago Yeah it's for simulating how many neutrons are going to be in the air.And what fraction of electromagnetic waves and particles are going to absorbed by those neutrons and other things in the air.Pretty much weather! idrissbellil 2 days ago *necular
dboreham 3 days ago I thought that's what "legacy weather simulation code" was! fooker 2 days ago Yeah it's for simulating how many neutrons are going to be in the air.And what fraction of electromagnetic waves and particles are going to absorbed by those neutrons and other things in the air.Pretty much weather!
fooker 2 days ago Yeah it's for simulating how many neutrons are going to be in the air.And what fraction of electromagnetic waves and particles are going to absorbed by those neutrons and other things in the air.Pretty much weather!
daniyal22 2 days ago Validating the numerical correctness across edge cases on 250k lines must have taken twice as long as the actual porting itself. rwallace 2 days ago Yes, I think that's what the paper is saying. The validation is the hard part, but the part that provides the value, because it's what makes the result actually useful.
rwallace 2 days ago Yes, I think that's what the paper is saying. The validation is the hard part, but the part that provides the value, because it's what makes the result actually useful.
Oh man I have to try this on all the legacy Fortran nuclear reactor simulator codes originally written in the 1960s
I thought that's what "legacy weather simulation code" was!
Yeah it's for simulating how many neutrons are going to be in the air.
And what fraction of electromagnetic waves and particles are going to absorbed by those neutrons and other things in the air.
Pretty much weather!
*necular
Validating the numerical correctness across edge cases on 250k lines must have taken twice as long as the actual porting itself.
Yes, I think that's what the paper is saying. The validation is the hard part, but the part that provides the value, because it's what makes the result actually useful.
The plagiarism machine is so good at this