The phys.org article and headlines are misleading, the authors did not investigate systems to actually transmit torque. From what I gather, the interesting findings are the parameters for co-rotation and counterrotation of the driving and driven cylinder, depending on the Reynolds number, distance and so on. To illustrate one of the images of their publication: https://i.imgur.com/m8P2iVw.png
That would make much more sense then what the article seems to imply (scientists reinvent a 100 year old torque converter! But worse!!). Of course that headline isn't nearly as fun (scientists develop better model for fluid dynamics in torque converter).
(Can't tell what the submission title was, but the byline (with quotes) works too: “Fluid gears” invention offers promise for improving mechanical devices)
Oh nice, that is a neat trick! One small nitpick (that makes no difference): The side lengths of the ISO Ax formats are rounded to the next mm, so actually the A0-format has an area of 0.999949m^2
I always enjoyed the "one-stop" solution with conda/mamba that installed the right version of cudatoolkit along with pytorch. How do you handle that without conda? (I'm genuinely ask because I never had to actually care about it.) If I manually install it, it looks like it is going to be a mess if I have to juggle multiple versions.