HGST Announces New Helium-Filled Hard Disk Drive Platform

Page 2 - Seeking answers? Join the Tom's Hardware community: where nearly two million members share solutions and discuss the latest tech.
Toggle sidebar Toggle sidebar
Status
Not open for further replies.
Isn't there a global shortage of helium?!

I know enterprise is a completely different beast.. but with the rate that SSD is improving, I feel like this is a technology that will not catch on.
 


It's unlikely that SSDs will be anywhere near HDD prices per GB any time soon, so improving HDD tech, especially in ways that can improve capacity per unit, can still be important for those who need a lot of capacity.
 


Many server cases don't have 5.25" drive bays, so the target audience would be very limited in the enterprise markets. Increasing platter count within the current 2.5" and 3.5" drive bays doesn't have this issue. For those who need great capacity in current cases/enclosures, this is a possible solution for improvement. For consumers, for the reasoning that you gave, it is arguably less useful.
 
[citation][nom]nukemaster[/nom]One word.....hindenburg[/citation]

Now that would be some big HDD 🙂


But seriously, the drive is so small that the amount of Hydrogen would not cause a major event, if any.

As for puncuring it/explosion...
again, the amount of gas and the size and structural strength of the -metal- case would not make this an issue. Prob. nothing more than a short sucking sound as air rushes in; the end.

 
[citation][nom]freggo[/nom]Now that would be some big HDD 🙂But seriously, the drive is so small that the amount of Hydrogen would not cause a major event, if any.As for puncuring it/explosion...again, the amount of gas and the size and structural strength of the -metal- case would not make this an issue. Prob. nothing more than a short sucking sound as air rushes in; the end.[/citation]

Puncturing a hydrogen-filled drive could be a fire hazard. An explosion seems unlikely, but it could have a short-lived spout of flame if the puncture causes a spark and the spark ignites the hydrogen.
 


Vacuum does not conduct thermal energy very well. So while it reduces friction significantly, the bearings, head actuator, read/write operation, other electrical components and platter rotation drive motor all generate heat that needs to be dissipated. In addition the the above mentioned gas buffer for the heads to float above the platter surface.

Now you may already know this and can imagine a design that is efficient with a vacuum in mass production then you should patent it.
 
[citation][nom]blazorthon[/nom]It's unlikely that SSDs will be anywhere near HDD prices per GB any time soon, so improving HDD tech, especially in ways that can improve capacity per unit, can still be important for those who need a lot of capacity.[/citation]
I guess you haven't seen ssd prices lately, prices are really coming down. Two days ago, I saw a 120 gb ssd at the egg for $60 bucks.
 
[citation][nom]cobra5000[/nom]I guess you haven't seen ssd prices lately, prices are really coming down. Two days ago, I saw a 120 gb ssd at the egg for $60 bucks.[/citation]

I can get a 500GB HDD for that price. Some 1TBs drop into that range occasionally. My point still stands.
 
[citation][nom]blazorthon[/nom]Many server cases don't have 5.25" drive bays, so the target audience would be very limited in the enterprise markets. Increasing platter count within the current 2.5" and 3.5" drive bays doesn't have this issue. For those who need great capacity in current cases/enclosures, this is a possible solution for improvement. For consumers, for the reasoning that you gave, it is arguably less useful.[/citation]

was looking at it from a consumer mind... but you make a point, servers wont have the space.

however we make a 5.25 drive, they get going on consumer and dont die off, well... there you go, they are around long enough to now make it a viable option to make a server version.

and its also a point to note that they will always have more space than a normal hdd, possibly allowing more space cheaper than needing 2hdds, and from what i know about servers, which isnt much, to make a drive hot swapable, requires a bit more space, so it may also take up less space too.
 
[citation][nom]danwat1234[/nom]I've heard that having helium can result in cooling problems because air helps to cool internal hot spots. I could be wrong though.Make a 5-platter 5TB 7200RPM regular hard drive already![/citation]
I assume helium has similar thermodynamic properties to air.
 
[citation][nom]blazorthon[/nom]I can get a 500GB HDD for that price. Some 1TBs drop into that range occasionally. My point still stands.[/citation]
You cannot compare technologies and only consider one factor. SSDs use less power, create less heat, are perfectly silent, and are way faster. They have no vibration and work in smaller spaces under more rigorous conditions. But yeah, your cheap $60 terabyte drive was cheaper.
 
[citation][nom]nukemaster[/nom]One word.....hindenburg[/citation]

Pure hydrogen (or any other pure element) cannot ignite or react without other gases/elements. It is as harmless as helium in this case and is actually better than helium since it is the lightest element (half the weight of helium), allowing for even lower densities. Hydrogen is also way more abundant (the most abundant element in the universe in fact).
 
[citation][nom]husker[/nom]You cannot compare technologies and only consider one factor. SSDs use less power, create less heat, are perfectly silent, and are way faster. They have no vibration and work in smaller spaces under more rigorous conditions. But yeah, your cheap $60 terabyte drive was cheaper.[/citation]

Yes, you can. When the purpose of these new drives is clearly to maximize storage density, then obviously price is a major factor. Most enterprise applications do not require the random access speed of an SSD. The cheapest (and only) non-PCI-E 1TB SSD on NewEgg is $2300. For less than $300 more, you can get a 20 pack of 1TB Toshiba SCSI drives.

So the price comparison for 20TB of storage is under $2600 vs $46,000. For 99.9999% of clients, the debate ends there as the benefits you listed for SSD's are practically irrelevant with that large a cost descrepancy. And that's just comparing 1TB. As you go even higher to 3 and 4 TB's, you're looking at $10k+ for one SSD vs around $400 for a traditional hard drive. For mass storage, SSD's are not an option for pretty much anyone simply because they cost too much. One factor.
 
[citation][nom]husker[/nom]You cannot compare technologies and only consider one factor. SSDs use less power, create less heat, are perfectly silent, and are way faster. They have no vibration and work in smaller spaces under more rigorous conditions. But yeah, your cheap $60 terabyte drive was cheaper.[/citation]

None of that matters when you need capacity at a price and don't need huge random performance, so your argument is moot, especially in highly storage-write intensive workloads where even the NAND flash, usually the most reliable part of an SSD, simply can't keep up unless you buy extremely expensive eMLC or even worse, SLC SSDs.

Your entire argument is irrelevant when a company needs four to sixteen terabytes or more of data storage per server and SSDs simply can't provide that much except like ten to twenty times more expensive (enterprise SSD prices versus enterprise hard drives is often even worse than consumer comparisons, especially since they don't get deals like having drives almsot half-off so often).

You're argument is like trying to get a company to use high-end Intel CPUs in a web server instead of an AMD Opteron that has higher parallel performance (the kind that matters in such an environment) at any given price. Sure, the Intel Xeons would be far faster for lightly threaded tasks, but that's not what servers are for so it's a huge waste of money on something that doesn't matter.

This article's technology improves the tech that many storage capacity-intensive servers and workstations need, so it's beneficial where it matters. Not everyone has the same needs as you, so basing your opinions on your own needs and then trying to apply your needs to other uses simply does not work at all.
 


The point is that it could be far more dangerous if the hard drive is damaged. If a helium filled drive's seal is broken, then you get a puff of helium in the room and a failed drive. RAID 5/6/10/whatever means that that doesn't matter too much. A hydrogen-filled drive can be more dangerous if it gets damaged.
 
[citation][nom]merikafyeah[/nom]Pure hydrogen (or any other pure element) cannot ignite or react without other gases/elements. It is as harmless as helium in this case and is actually better than helium since it is the lightest element (half the weight of helium), allowing for even lower densities. Hydrogen is also way more abundant (the most abundant element in the universe in fact).[/citation]


Helium and hydrogen have the same effective mass (hydrogen is diatomic). The advantage of helium is it isn't as reactive or explosive (when exposed to air). The con is that the world supply of helium is almost gone. I agree though, hydrogen would be the better choice!
 
[citation][nom]whitecrowro[/nom]vacuum is better[/citation]As Otaku pointed out, hard drives don't work in a vacuum (unless they're sealed with gas inside). The heads require air pressure, they "float" (air acts as a bearing) at a specified height above the platters. Insufficient pressure would cause head crashes, actual vacuum would mean the heads just drag along the platters, decimating the drive in no time. So no, vacuum is not better, and you shouldn't have assumed that their engineers are stupider than you. Seriously.
 
[citation][nom]husker[/nom]You cannot compare technologies and only consider one factor. SSDs use less power, create less heat, are perfectly silent, and are way faster. They have no vibration and work in smaller spaces under more rigorous conditions. But yeah, your cheap $60 terabyte drive was cheaper.[/citation]

It's all about what your doing with them as Blazorthon pointed at. Not everything shown on the web site is pointed towards the people that do general things with there computers.

This article more points towards the Enterprise folks that need the best combo of Overall storage space, power management, low heat, and cost. SSD fit the bill in terms of power and heat. Although thinking storage and cost is still way out there compared to HDD.

Not to say the SSD haven't come down in price but as others pointed out, when you start thinking 10 or or more drives that has to meet a capacity quota, ssd is just not worth the money.
 
I design cryogenic, ultra high vacuum, low-particle generating devices. So let me get a few facts out for you guys.

The fluid in a hard drive enclosure whether it be air (roughly 80% N2 and 20%O2) acts as a hydrodynamic bearing, keeping the head from touching the platter. The fluid is also a damper, reducing vibrations. Third, the fluid transfers heat via convection - the most efficient form of heat transfer. While liquid helium is a poor thermal conductor, the gas is actually much better than gaseous nitrogen, whereas liquid nitrogen is very good at extracting heat compared to liquid helium. So switching from air to helium will improve thermal performance over air.

About running out of H2: Yes it's a limited resource, but we aren't running out any time soon. It is extracted during fracking of natural gas of which there is no shortage. However, it is expensive to do so, therefore many gas companies simply do not want to do it, which is why we are seeing H2 prices go up. (Helium refrigerators are not cheap - on the order of several million dollars for several hundred watts of capacity) Currently liquid helium costs about $7/liter. It has an 800X expansion ratio at 1 atmosphere. So while H2 in liquid form is expensive for cryogenic applications, the gas isn't nearly as bad, and would barely add any cost to a HD.

Could you make an HD work in vacuum? Probably, peizo motors (heads) do work even in a UHV environment, but there are other problems, notably the bearings. There are vacuum greases out there, and dry lube options, but they simply don't work as well as in atmosphere. Dry lubed bearings (whether it be WS2 or MoS2) also generate many submicron particles. And heat can only be transferred via conduction or radiation in vacuum. No convection occurs without a transfer fluid. Vacuum could work, but there are many technical hurdles to overcome, which often are not cheap.
 
Here's a slide from WD HGST product presentation. http://www.extremetech.com/wp-content/uploads/2013/11/HGST_6TB_HE6HDD_comparison.jpg

They show a side by side of the "old" & new He drive... with gasket seals... ROTFFLMFAO. The He isn't the invention. The invention is the sealing technology's: yield at 100% or at least 99.9%, fabrication cost of sealing, field reliability, etc. Electro-resistive seam sealing/welding was an old way of doing it.
 
Status
Not open for further replies.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom