Computers

HGST's helium-filled HDD offers a world-first 10 TB of storage

HGST's helium-filled HDD offer...
HGST's Ultrastar Archive Ha10 is aimed at enterprise users
HGST's Ultrastar Archive Ha10 is aimed at enterprise users
View 3 Images
HGST's Ultrastar Archive Ha10 uses shingled magnetic recording (SMR), a hard drive technology that records data on overlapping rather than parallel tracks
1/3
HGST's Ultrastar Archive Ha10 uses shingled magnetic recording (SMR), a hard drive technology that records data on overlapping rather than parallel tracks
To begin with, the Ultrastar Archive Ha10 will be available to cloud and OEM storage clients
2/3
To begin with, the Ultrastar Archive Ha10 will be available to cloud and OEM storage clients
HGST's Ultrastar Archive Ha10 is aimed at enterprise users
3/3
HGST's Ultrastar Archive Ha10 is aimed at enterprise users
View gallery - 3 images

We first caught wind of HGST's high capacity hard drives in 2012, when the company claimed it could boost storage capacities by 40 percent by replacing regular old air inside the drive enclosure with helium. The Western Digital subsidiary stayed the course, producing a helium-based 6 TB HDD in 2013 and 8 TB model in 2014, and has now continued the upward trend with the world's first 10 TB hard drive.

Aimed squarely at enterprise and data centers, the Ultrastar Archive Ha10 is the company's latest take on a helium-based HDD. The apparent benefits of helium when it comes to HDD storage is its markedly lower density of around one-seventh that of regular air. This means less friction with internal moving parts, resulting in less power needed to drive the device and increased data density of the individual disks.

HGST calls its version of this HelioSeal and has combined it with shingled magnetic recording (SMR), a hard drive technology that records data on overlapping rather than parallel tracks, much like roof shingles (hence the name). The company says this results in an industry-leading storage density, low power consumption and ever-reliable storage solution. However, since SMR requires the writing of entire tracks, the drive is suited for active archive duties rather than frequent update workloads.

Further to its mammoth storage capacity, the 3.5-in drive is rated at two million hours mean time between failure (MTBF) and a 10-15 unrecoverable reduced bit error rate with 600k load-unload cycles. It comes in SATA 6 Gbps and SAS Gbps varieties and HGST claims a 20 percent improvement inWatts per TB over the preceding Ultrastar He8.

To begin with, the Ultrastar Archive Ha10 will be available to cloud and OEM storage clients with the software capabilities to harness the density of the device.

Source: HGST

View gallery - 3 images
10 comments
Chris White
They problem with this is that there is limited quantity of helium, and the cost of it is kept artificially low. Much He is going to waste supporting the party balloon industry. Now we are going to waste it with the computer industry. Use of what in reality is a valuable and precious gas should be reserved only for those industries where it is absolutely necessary and vital. We need to stop using it for non-critical applications.
4Freedom
Agreed Chris, I was going to say the same thing. In light of the helium shortages that are affecting medical scanning machines, this seems quite unnecessary.
BenC
As size per drive goes up and transfer speeds to/from the drive remain constant, it becomes harder/slower to get at the data on that disk.
Douglas Bennett Rogers
Product of nuclear fusion but at very low rate.
mystixa
While we do need to stop wasting helium these old cliche examples are not valid. Party balloon helium is recycled helium with massive amounts of contaminant gas, almost exclusively from MRIs which need to change out their helium supply on a regular basis.
How about going after all the lighter then air craft that uses massive quantities of the stuff. Beyond that why not fill weather balloons with hydrogen instead? The hindenberg excuse cant be used for those when they carry disposable instruments up everyday anyway.
DonGateley
I rather doubt that the volume of Helium used in one of these times the number of devices that may ultimately use it makes much of a dent in supply. That is merely a guess, however.
Helium has a nasty habit of finding it's way out of things that try mightily to enclose it for any length of time and I wonder what their solution to that is.
EH
The amount of helium used in a hard disk is negligible compared to a single party balloon - we're talking a few cents of helium in a device that costs hundreds of dollars.

What I think is exiting is that, unlike older drives that have breather ports to equalize pressure between the inside and outside of the drive, this is a sealed unit. That means there shouldn't be any altitude limit for operating the drive, which has been a problem in observatories and other high-altitude areas, as well as in aerospace.
EH
I checked how much helium is used - HGST says 10,000 drives per tank of helium. The biggest standard tank is 300 cubic feet, so that's 0.03 cu. ft. In contrast, an 11in. balloon uses 0.53 cubic feet, nearly 18 times as much. (1 cu. ft. = 28.3 liters, 0.03cu. ft. = 0.85 liter)
Craig Jennings
I wonder how they keep the helium in there? It would be fantastically sealed, at low pressure I suppose they don't need to worry about diffusion.
artmez
Ignoring the issue of Helium (He) containment which was already commented upon, so WHY was Helium chosen over air? Or H2 (are He atoms smaller than H2 molecules?)? Is it He's inertness (i.e. it's not an oxidizer)? Or that He's thin and has less fluidic resistance between the flying head and disk surface?
It is the presence of O2 as an oxidizer that's bad? I guess that depends on the drive's oxide coating -- is it iron, cobalt, an alloy, or ferrite ceramic? Iron oxide comes in two flavors: ferric (+3) and ferrous (+2), and the later can oxidize to the former. I don't think it's the oxidizing properties that are the main reason He is used. Otherwise dry N2 would be use like in tires.
If it is oxidation and "wind" resistance, then why not use a vacuum which won't cause either? The problem with the vacuum is that an HD needs fluid between its "flying" head and disk surface to prevent damage to both. I'm guessing that since He is the smallest inert gas, it lets the disk ride closest to the disk surface than any of the other inert gases, reducing the amount of energy needed to flip magnetic poles at the disk surface and also has the smallest magnetic pole size due to the reduced smearing of the magnetic field's lines of flux.
If the seals effectively contain the Helium, then that will also prevent air from getting inside since the molecules of N2 and O2 are so much larger than He atoms. And if He does escape, then that would create a vacuum inside the HD relative to the outside air pressure. So how do they seal it in?
The older articles claims the hermetic seals allow liquid cooling for enhanced electrical efficiency. I'll have to think about that one some more. Maybe that helps to seal the Helium in better?