News Tuolumne is El Capitan's Little Brother with 200+ PetaFLOPS Performance

[Admin]

Lawrence Livermore National Laboratory preps unclassified system with AMD's Instinct MI300 APUs.

Tuolumne is El Capitan's Little Brother with 200+ PetaFLOPS Performance : Read more

 

[NeoMorpheus]

Another awesome win for AMD!.

 

[bit_user]

El Capitan is expected to be used primarily for stockpile stewardship, the U.S. program of nuclear weapon reliability testing and maintenance without the use of actual testing.

I find it a little amusing that we've been hearing this same justification of the DoE's supercomputers for a few decades, at least. You'd think that if the first few generations of such machines were capable of answering these questions, a modern desktop with a decent dGPU should be all it takes in this day and age.

That's probably a very simplistic and ignorant take. If anyone can enlighten us on why 2 ExaFLOPS of compute power is needed for nuclear stockpile stewardship, please be my guest.

BTW, I have no doubt the big machine will get plenty of good use. I just wish they'd drop their former pretense, if that's all it is.

 

[Allen_B]

I find it a little amusing that we've been hearing this same justification of the DoE's supercomputers for a few decades, at least. You'd think that if the first few generations of such machines were capable of answering these questions, a modern desktop with a decent dGPU should be all it takes in this day and age.

That's probably a very simplistic and ignorant take. If anyone can enlighten us on why 2 ExaFLOPS of compute power is needed for nuclear stockpile stewardship, please be my guest.

BTW, I have no doubt the big machine will get plenty of good use. I just wish they'd drop their former pretense, if that's all it is.

That seems like a fair question. By now we ought to be able to be fairly confident that existing warheads will go bang! if ever called upon.

One guess for why we might need so much additional compute power is to design new weapons. There has always been a desire to reduce size and weight, as well as reducing the amount of fissile material needed. Plus other optimizations such as reducing radioactive fallout, edging closer to a pure fusion design, etc. There's probably plenty of work to go around.