3 ms·
But that's current hardware. What about future hardware? What about hardware optimized for inference? What about hardware optimized to run a particular model?
by sanderjd 4mo ago
But that's current hardware. What about future hardware? What about hardware optimized for inference? What about hardware optimized to run a particular model?