106 Gbps
They get to this result on 0.6 MB of data (paper, page 5)
They even say:
Moreover, there is no need to evaluate our design with datasets larger than the ones we have used; we achieve steady state performance with our datasets
This requires an explanation. I do see the need - if you promise 100Gbps you need to process at least a few Tbs.
A few years ago people were talking about convergence of phone/desktop, i.e. you plug your phone into a big screen and keyboard and it’s now your desktop computer.
Mobile apps are shit for that. Sure, my phone is powerful enough to browse internet, play video and music but on desktop with mouse/kb it’s just weird and funky. And I’m not even talking about any productivity software which is straight impossible.
Another self portrait, drawn when he was 90 or 91. Probably my favorite of his self portraits. Titled “The Young Painter”:
It was incredible to see it live unprepared. When you look chronologically through his paintings, you see basically every modern style there is - the guy participated in a lot of art movements over the twentieth century—and was proficient and productive in several of them. He starts classically, but soon descends into surrealistic nightmares and all the other things he became famous for. And then, finally, in the end, after all this insanity of lines and cubes and shapes and trying to figure out meanings (or at least subjects), you come to the last painting in the exhibition, and it really looks like something a talented ten-year-old could draw - full of life and innocence and optimism.
Based on your analogue they drive the car for 7.5 inches (614.4 Kb by 63360 inches by 20 divided by 103179878.4 Kb) and promise based on that that car travels 20mph which might be true, yes, but the scale disproportion is too considerable to not require tests. This is not maths, this is a real physical device - how would it would behave on larger real data remains to be seen.