I am an author of one of the earliest RAID papers, I also funded the UC Berkeley group ( Garth Gibson, Dave Patterson, etc). First off the original paper on RAID described a "Taxonomy" of RAID arrays and came up with the formalized math to describe them and their properties. RAID disk arrays had been around since the early 1970s.
Second... the thing about HDDs compared to other aspects of computer systems was / is that they are electro-mechanical devices... i.e. a motor spins a platter of rust under a magnet to either store or retrieve your data (usually more than 1 platter). The magnet must also move from the outer rim to the inner rim. Storage capacity has improved over the decades as a result of technical improvements in the size of the magnet, the surface of the rust, and the distance the magnet (head) is over the platter. But the speed of rotation and speed of movement of the head across the surface of the platter... thus the "seek time" has hardly improved since the 1970s. Not when compared to the speed of electronics of the CPU and memory systems or even networking speeds.
The primary reason for RAID was not creating redundancy of data... it was to increase the speed of data transfer.
If I spread your file across, say, 8 drives... I can potentially read and write it 8 times faster. That said... I now have to worry a little bit that should I lose even 1 drive... I have lost all the data because it will be missing 1/8th of the data... hence the second requirement or including some sort of data protection (RAID 1,2,3,4,5,6,10 but NOT RAID 0). Each of the RAID levels above 0 describe a different method to create and save enough redundant information about the related data that should 1 (or possibly more) drives fail, the missing data can be recreated.
His wind turbine wall has none of these characteristics other than it could be imagined as a 2 dimensional array of wind turbines.
Doesn't mean it is a bad idea... just don't compare to computer RAID arrays.
https://www.computerhistory.org/storageengine/u-c-berkeley-paper-catalyses-interest-in-raid/
= new reply since forum marked as read
