That is good to know. Is 12 TB as of 2024 the smallest capacity that is guaranteed to be a helium-filled drive among all manufacturers?
Such information should be disclosed as it pretty much changes how the drive behaves as far as I am concerned. But then again those same manufacturers felt that the difference between CMR and SMR drives didn’t need to be disclosed. At least the presence of helium can be inferred from other specs (noise, power consumption, presence of screw holes in the middle of the drive).
SMR isn’t actually bad, at least for archiving purposes, but the fact some manufacturers hide this information from customers is unforgettable imho, as it’s directly related to performance and also driver compatibility.
As for helium, yes we can relatively easily tell this, because helium exterior design is totally different. Having said that, I’m absolutely happy with my old air filled 8TB WD Plus drives, they are just fantastic.
While this is theoretically possible, bearing companies have been selling bearings with different service limits for decades now. This is a business with very fine tolerances and well-honed manufacturing processes that have to be extremely consistent.
The key issue with binning is understanding how to predict different service lives in advance. With chips, this is somewhat easier since one can infer quite a bit of information from how they perform, what cores show up under what voltages / frequencies, etc.
Doing the same thing quickly in a HDD factory seems possible but unlikely. Not saying it cannot be done (my colleagues did a lot of that kind of optimization work across multiple industries around the world) but getting to primary and secondary process variables to start predicting future performance of a mechanical device can be tricky at best.
