iPhone 17 Pro Demonstrated Running a 400B LLM

Re:iPhone 17 Pro Demonstrated Running a 400B LLM by ottah on Monday March 23, 2026 @12:38PM (#47491830)

That's actually pretty cool, but I'd hate to freeze a models weights into silicon without having an incredibly specific and broad usecase.

[ Parent | on HN ]

Re:iPhone 17 Pro Demonstrated Running a 400B LLM by patapong on Monday March 23, 2026 @01:39PM (#47492644)

Depends on cost IMO - if I could buy a Kimi K2.5 chip for a couple of hundred dollars today I would probably do it.

[ Parent | on HN ]

Re:iPhone 17 Pro Demonstrated Running a 400B LLM by whatever1 on Monday March 23, 2026 @01:42PM (#47492689)

I mean if it was small enough to fit in an iPhone why not? Every year you would fabricate the new chip with the best model. They do it already with the camera pipeline chips.

[ Parent | on HN ]

Re:iPhone 17 Pro Demonstrated Running a 400B LLM by superxpro12 on Monday March 23, 2026 @01:46PM (#47492750)

Sounds like just the sort of thing FGPA's were made for.

The $$$ would probably make my eyes bleed tho.

[ Parent | on HN ]

Re:iPhone 17 Pro Demonstrated Running a 400B LLM by chrsw on Monday March 23, 2026 @02:04PM (#47493006)

Current FPGAs would have terrible performance. We need some new architecture combining ASIC LLM perf and sparse reconfiguration support maybe.

[ Parent | on HN ]

Re:iPhone 17 Pro Demonstrated Running a 400B LLM by 0x457 on Monday March 23, 2026 @02:49PM (#47493575)

Wouldn't it be the opposite of freezing weights?

[ Parent | on HN ]