Low flying heads
Interesting to see how the heads fly above the platters in vacuum ( filled? emptied?) drives. Electrostatic charges perhaps?
HGST has announced second-generation helium drive tech after shipping a million gen-1 Helium drives and upping field reliability by 15 per cent. These drives have enclosures filled with helium instead of air. This is only 14 per cent of the density of air, so platters spin with less friction, vibrating less and needing less …
We surely don't want to run out of helium but you might want to check your target: wild-ass-guess says one helium party balloon has the same He gas content as 100+ drives (since most of interior volume is platters, heads, etc), and the average wedding party around here seems to release a couple of dozen balloons, symbolising aspirations or somesuch. So that's a few thousand drives worth, each of which will probably be longer-lived than the average marriage ;-)
Maybe Tim Worstall will drop in and tell us what the real He reserves & resources are like.
Foil coated balloons (specifically made for helium filling and so far less pervious than standard rubber ones) steadily lose both buoyancy and volume across several weeks before entering an apparently endless half-sized afterlife (observations terminated after some months when child finally persuaded that balloon had joined choir invisible). Interestingly the foil is only a little elastic at the outset and rapidly becomes flaccidso most of the helium loss occurs at atmospheric pressure rather than being "squeezed out" by the taut rubber membrane. So from this I guess that the contents is roughly half He and half dull (air? nitrogen), i.e. indeed only half a metric crapton of new-fangled HDD per wedding.
I grant that having a medium to 'fly' the heads is important, but surely the major problem is not preventing the helium getting out, but preventing other atmospheric gasses getting in.
Because nitrogen, oxygen, argon and carbon dioxide are much larger molecules, it's much easier to prevent them entering the drive.
Provided you can prevent other gasses entering, the rate of helium leakage will fall anyway, as the pressure inside the drive drops, so I can see them having a quite long life.
Helium molecules are actually atoms. They are the smallest thing floating around out there. They ooze out through thin layers with relative ease. I'd imagine the pores in metal are pretty good at stopping it though.
Why the hell we waste He in weather balloons is beyond me - having spent time deliberately filling things with H2 and setting them off its bloody hard to hurt yourself doing it. Well unless you get some 02 and mix that in as well. That makes your ears ring for days!
I'm guessing that the reason there is not what happens to the balloon itself, but because driving to where you want to launch the thing with a tank of pressurised Hydrogen in your vehicle has a whole can of worms around permits, vehicle labelling, where you can and cannot go and such associated with it.
The article mentions that He10 is an SMR drive, but it doesn't explain what that means. SMR, particularly device managed devices like the He10, are not at all comparable to regular spinning disks.
Device managed SMR drives are only really suitable for write once/read many archival workloads, because of the massive expensive (performance) cost of the shingling on writes combined with the non-aware access of the device - the OS will treat it like any other type of drive - and you will get atrocious performance.
This review includes a comparison of rebuilding a RAID array comprised of 8TB Seagate SMR drives, compared to HGST He8 drives, which are PMR. The rebuild took 57 hours on the SMR, compared to 20 hours for the PMR drives. Average read/write speeds during the rebuild were both around 155MB/s on the PMR array and <10MB/s on the SMR array.
Host managed SMR drives (which are not out yet) will allow the OS to understand the performance characteristics of the drive, and use it in a totally different way to a PMR drive. Until then, they are really only useful in a single disk archive.
It really doesn't matter that shingled drives are only good for archiving. That's the role of all spinning drives in the future, as solid-state gets cheap and very dense.
Intel and others are predicting equal capacities next year, and price parity in 2017. when this happens, the spinners are doomed!
"Helium molecules" ? Ain't no such animal. Like all the noble gases, helium in its normal state exists as isolated atoms, both lighter and smaller in size than the molecules of gases such as oxygen or nitrogen. That's why helium leaks out so well, since the atoms can get through the tiniest of pores.
I'm surprised the manufacturers didn't give Neon a shot. Cheaper all around and similar properties.
Neon is much much more expensive than Helium. So while Neon does not when released into the atmosphere escape into outer space as Helium does, and thus not be lost to humanity, it is more expensive. That said it probably costs a few pennies for the amount required to fill a drive.
This post has been deleted by its author
It doesn't leave a vacuum behind!
It will only continue to escape if there's a pressure differential. So leaking helium is either replaced by air or (maybe more likely) outgassing from internal components. That would leave a denser, more frictiony gas which may still cause problems as the whole set up is dependent on working in a low friction environmnet.
And 'frictiony' is so a word.
It so does leave a vacuum* behind!
A sealed, air filled, hard disc drive will always have approximately the same pressure as the ambient air because the rate of inward diffusion matches the outward rate.
In the case of a sealed, helium filled, drive there is almost no helium outside, but the helium atoms inside will move at a velocity determined by the ambient temperature and their random walks will lead some helium atoms outside. (See: Kinetic Theory of Gases). Far more Helium atoms will move out, than air molecules move in, because the air molecules are larger and are less likely to 'find' paths through the seal. I wouldn't expect a spectacular drop in pressure in 5 years, but I bet you could measure it with an unsophisticated manometer.
*partial vacuum - to be precise. It's never going to reach a high vacuum!
Mine's the one with the 'diffusion pump' in the long pocket.