Impressive work! For comparison, an H100 can do roughly 100-200 tok/s of Muse 30B for raw inference single-thread, going up to low thousands of tok/s with a large number of threads - and I am sure that for the massively-multi-threaded case they can optimize the prover further. So we roughly, sort of, have single-digit (<10x) overhead for LLM proving! Next step is getting single-digit overheads for FHE, and then ultimately vFHE (aka STARK * FHE). A crazy ambitious milestone given present FHE overheads, but because of how highly structured and almost-linear LLM inference is, it's closer to the realm of possibility than you might think. Single-digit-overhead all the things.

Again, I will urgently repeat: The biggest fascistic move the Trump administration is making is destroying Public Health. They have ensured we will be living through an Age of Pestilence. The People must care for ourselves!

Mayor Mamdani, @NYCMayor What will you be doing to protect the children of New York from Donald Trump and Robert Kennedy Jr's attack on Public Health and their intentional weakening of the school vaccine mandates, and making obtaining vaccines extremely difficult, while maximizing the religious exemptions to school mandates? A lot of children will die from unnecessary illnesses!

Just to reiterate: the Vaccine Miscarriage story is a distraction from Trump and RFK Jr's gutting of school vaccine mandates yesterday. They are literally sacrificing children in honor of the Regime. They are MOLOCH WORSHIPPERS! WAKE UP WAKE UP WAKE UP WAKE UP WAKE UP WAKE UP!