@Lerg @ZackKorman @WhiteHouse Would kill for $500 a year. Have multiple customers who have "onprem" LLMs running on managed GPUs. And yet if one was t

@Lerg @ZackKorman @WhiteHouse Would kill for $500 a year. Have multiple customers who have "onprem" LLMs running on managed GPUs.

And yet if one was to encrypt the server in FIPS mode, suddenly everything breaks because of needed python libraries.

Almost like folks were kinda stretching meaning of onprem.