Ritvik Rawat
@rawat_ritvik · 10 nov. 2025
The other arguably more important factor (given 4b anything is just fine for neural nets) is the ability to compose smaller muladd units to create larger data types.
Elon Musk
@elonmusk
🎯
01:27 · 11 novembre 2025 · 17,9 k vues
33
26
351