[exikyut]

You might be able to get partway there by setting up thunderbolt networking between two endpoints. The caveat is that USB 4.x is still new enough it may require a significant time sink to get working at ~40Gbps et al, so your testing wouldn't be able to scale that high. But you can go to 10Gbps without too much effort: https://news.ycombinator.com/item?id=39003469

USB is quite the rabbithole, with a lot of functionality often hidden: https://news.ycombinator.com/item?id=39470381 - my takeaway from this is that "the PHY in the controller is 99% of the SERDES you wish you had, you just have to figure out how to make it cooperate"

The chipsets found attached to ARM devices and SoCs etc may be easier to poke around at, but they are likely to either be expensive or just USB 3.0. :)

Then there are devices like the FNB58 and the Power-Z KM003C that interrogate the controller chip in the cable and tell what charging protocols the cable supports, and also measure practical charging voltage and current (and you can acid test chargers using USB charger dummy loads). These devices don't test the cable's speed though. (On that note, https://old.reddit.com/r/UsbCHardware/comments/1b5wgc6/how_t... - that subreddit may also be interesting)