4 ms·

I don't think you are correct on this, with ML workloads becoming a thing, increasing speeds across the board is extremely important.

• pdimitar 3y ago
  And? You're saying all of us will train models all day?