Elon -
My hope was to help align everyone's vision for SpaceX making life multi-planetary by releasing this interview right after the IPO on Friday.
We tried incredibly hard to finish the set build last month but the space suits took longer than expected to get right.
I know this week is likely the most jammed you will ever be and it would kill me if we can't make this happen.
With that in mind, I'm prepared to greenlight whatever spend is required to keep the set up until you're able to make it.
Just need a date.
Lmk
Ti
SpaceX has almost finished writing V1.0 of an in-house AI training stack in C that exact-maps to 220k GB300s with 800G NICs, making heavy use of pipeline parallelism and getting as close to bare metal as possible.
The potential speed improvement vs JAX for large training runs is over an order of magnitude.
@BrontoM305@bfmbusiness , elle devrait faire 416 jours, on sera donc, si jamais un rythme de 10k/an, à plus de 1 par heure ( + de 1 par heure si c’était pas assez clair)
2/2
@BrontoM305@bfmbusiness Toi t’es un champion. Ce rythme est attendu d’ici 5ans, donc c’est dans 5 années que ce nombre de lancements sera attendu. Ensuite justement, 10000 par 24 fait 416, donc ton calcul démontre justement que si on lançait une fusée par heure durant une année 1/2
Onboard views from Starship and Super Heavy V3, which are equipped with upgraded cameras capable of streaming 4K video through every phase of flight via @Starlink
@Tipolo91@le_Parisien suivant le projet depuis plusieurs années, je trouve le terme "globalement" assez approprié. Le vol s'est déroulé mieux que je ne le craignais, moins bien que ce que j'espérais.