CERN uses tiny AI models burned into silicon for real-time LHC data filtering

Re:CERN uses tiny AI models burned into silicon for real-tim... by moffkalast on Saturday March 28, 2026 @07:21PM (#47558969)

I'm borderline shocked that all of this extra overhead is somehow more efficient than something as simple as computing both branches or something.

[ Parent | on HN ]

Re:CERN uses tiny AI models burned into silicon for real-tim... by ahartmetz on Sunday March 29, 2026 @04:28AM (#47561352)

The required computing resources double at every branch where you take both paths, and if you speculate ahead by 100+ instructions, with let's say up to 20 branches, it gets way out of hand.

I could see CPUs sometimes taking both paths for close, hard to predict branches. Does anyone have information on that?

[ Parent | on HN ]

Re:CERN uses tiny AI models burned into silicon for real-tim... by tyjo99 on Monday March 30, 2026 @01:12PM (#47577007)

I believe the way things are currently trending is that architectures might turn some short hard to predict branches into predicated instructions instead (similar to x86 CMOV or some ARM conditional execution instructions). Outside of short branches the overhead for loading up to 2 instructions for every 1 that gets executed can be too costly. Branch predication on SIMD/SIMT instructions is already the way things work for GPUs and AVX256/512 from my understanding.

[ Parent | on HN ]