@tbpn Seeing this + sams response to reinstating jensen makes me think its pretty likely this was coordinated. Tbpn is an openai company. Anthropic also has people significantly involved if you go look at the details. Seems like closed source really hates jensen
@elonmusk@cradrrat Whose idea was this? The overhead a new stack on this level is high. I think just avoiding codegen and hillclimbing would get very close to this perf and help move faster than a new stack purely from engineering standpoint, but maybe this new stack benefits AI coding somehow??
i don't like this argument. there's clearly enough food, medicine, energy for all humans on earth. sure, we can cure cancer and whatnot, but why assume these problems are unsolvable bc lack of 'intelligence' and not a lack of real world wherewithal to fix them?
@bryancsk Someones lying here lol how do the agents take that long if they use levent and tristans work. even if its in training data i feel like thats too many agents and time
By now youโre already familiar with MTIA, our family of in-house training and inference accelerators.
Today, weโre sharing a deep dive into MTIA 300, the first MTIA chip optimized for training ranking and recommendation models. MTIA 300โs built-in NIC chiplets allow it to meet the communication needs associated with training recommendation models with superior performance over general-purpose GPUs.
See the full technical deep dive from the @ Scale: Networking Conference: https://t.co/ajpH6s5Ryr
@Mononofu@JensenHuang Supporting open source != open sourcing. You are MTS, use your brain and stop letting claude do all of your thinking once in a while