• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

When will the 5TB drives come out?

Page 2 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
Wrong. I have a snapshot RAID setup at home, and I use nothing but 5400rpm drives. When I lose a drive, I have to completely read the remaining drives, and completely write to the replacement drive (well, technically, not all of the drives get completely read if the lost drive is less than 4TB, since some are 2TB, 3TB, 4TB...but the replacement drive is completely written).

Also, most cloud storage businesses use 5400rpm drives in the RAIDs. They are going for large capacity, low power usage, not speed. There are a huge number of HDDs used in cloud storage, so that is a big chunk of the market. Real world, indeed.

Good for you.

But most people looking at these capacities want it for media, and again, they don't RAID (because they think it's too complicated).
 
But most people looking at these capacities want it for media, and again, they don't RAID (because they think it's too complicated).

I'm not sure why you quoted me but did not read what you quoted. You apparently missed the part about cloud storage.

Also, the point I made applies to any sort of redundancy, including mirroring, WHS duplication, or even manual backups of drives. Whenever a drive fails, the replacement drive gets mostly or completely rewritten, and the faster the better.

Yes, there are some people who buy 4TB drives and have no redundancy or backup of any kind. But it is a huge stretch to claim that most 4TB drives are used without any such redundancy.

And as I said before, cloud storage businesses are huge consumers of large, low-power HDDs. And they definitely have some redundancy.
 
Last edited:
At 150MB/s it only takes 33 mins to fill the entire drive with 5TB of data.

You may want to crunch those numbers again. EDIT: I see someone already corrected you. It bears mentioning that as platter data density rises then holding spindle speed constant you would expect higher MB/s sequential transfers, but it's still not exactly fast.

you are correct. But please try not being so rude and condescending.

He's on my ignore list for a reason.

What makes 5 platter drives less reliable? Heat? I realize that there are more read heads, is the issue that it is more likely to have a head crash?

More moving parts = more potential failure points.

Good for you.

But most people looking at these capacities want it for media, and again, they don't RAID (because they think it's too complicated).

RAID-style setups will become commonplace as Windows Storage Spaces becomes commonplace. About time, too, given the failure rates of HDDs. 5xxx rpm is sufficient and imho preferable due to the power savings... we're already at the point where the fastest 5xxx rpm drives are getting close to saturating GigE connection speeds on average. With another boost to density, 5xxx rpm drives will be pushing at the GigE real-life transfer speed limit. So 7200rpm gives diminishing returns unless they are local drives.
 
Last edited:
I just found this interesting read. Lotsa good info there.

It seems to suggest 25% per year is the new standard/goal.
"The use of SMR and then HAMR will not change the approximate 25 per cent a year increase in area density. That seems set to be a fixed upper rate for a good few years."
 

I would be extremely leary of this. Of course maybe I am biased after being burned by the original Hitachi Deathstar when they went with 5 platters while everyone else limited to 3 in 2000.

A normal HDD has a hole for pressure equalization with the outside (with a filter keeping particles from going through the hole. Obviously if the drive is helium filled, you can't do this, so you are going to have the pressure inside the drive increase and decrease depending on temperature. Which will cause the metal casing to bend a bit.

Also, they conveniently forget to mention that at any given temperature, Helium atoms are moving ~ 2.7x faster than air molecules, which in addition to being smaller, makes them more likely to escape from a container if the seal isn't perfect (or escape *through* the container if there is enough space between molecules of it), which is why a helium balloon never stays full more than a day or two. Yes, I know the drive would have a better seal than a ballon, but still, do you trust the helium to stay in there for the 5 year drive life cycle?
 
Also, they conveniently forget to mention that at any given temperature, Helium atoms are moving ~ 2.7x faster than air molecules, which in addition to being smaller, makes them more likely to escape from a container if the seal isn't perfect (or escape *through* the container if there is enough space between molecules of it), which is why a helium balloon never stays full more than a day or two. Yes, I know the drive would have a better seal than a ballon, but still, do you trust the helium to stay in there for the 5 year drive life cycle?

The articles do state that the advantages of helium filled drives has been known for many years and the breakthrough was in manufacturing a seal that is both effective and cheap.
 
Yeah, it is. Guess I got my divisions wrong.

But nearly no one is gonna need to transfer/store 5TB to the HDD right away, so it's inaccurate/non-representative of real world use.

And a 1PB HDD would be well over 400 or even 500MB/s due to platter density, so that 77 day number is meaningless.

the RAID analogy is fine. some home users would have a media centre RAID system to have everything on 1 drive

also, would you trust 5TB of data on an untested drive? i always fill mine with junk data (truecrypt it) and do a full chkdsk for bad sectors. i've had a drive die on me during this so imagine if i was cut/pasting files from old to new and that happened?
 
I'm ready to pay a premium just to get rid of the noise of 3.5 inchers.

Do you guys sit with your ears planted on the side of your computer cases? Seriously, I have to struggle to hear my hard drives and I have a server about 8 feet away with six WD 750 GB blacks in an array and my main computer is a couple of feet away with a couple of spindles.

Back to the OP, I had heard about HAMR as well and thought that we'd see 6 TB drives within the next year or so, but who knows if that will actually happen since others seem to think Seagate can't deliver with this technology.
 
Do you guys sit with your ears planted on the side of your computer cases? Seriously, I have to struggle to hear my hard drives and I have a server about 8 feet away with six WD 750 GB blacks in an array and my main computer is a couple of feet away with a couple of spindles.

It is there, and it is easily noticeable if you have a secondary HDD (one the OS is not installed on) and let windows turn off HDDs after idleness.
 
Reality seems to be that 5 TB hard drives will not be released by Western Digital until late 2013, see:

http://www.xbitlabs.com/news/storag...veil_5TB_Hard_Drives_in_Late_2013_Slides.html

Also, according to the article WD will not introduce Green and Red 4 TB drives until the third quarter of 2013.

The joys of oligopoly... :\

Well darn, I was hoping a 4 TB Red was around the corner for my new server build. I've heard they're having reliability issues with their 3 TB Reds (anyone confirm?), so maybe it is better if they take their time and get the 4 TB models right.
 
It is there, and it is easily noticeable if you have a secondary HDD (one the OS is not installed on) and let windows turn off HDDs after idleness.

I do have secondary spindles. Heck, I even used to have a 1.5 TB Green drive and while I could hear it spin up after being idle, it was not so annoying that I'd want to eliminate 3.5" drives for 2.5" drives at this stage.
 
A normal HDD has a hole for pressure equalization with the outside (with a filter keeping particles from going through the hole. Obviously if the drive is helium filled, you can't do this, so you are going to have the pressure inside the drive increase and decrease depending on temperature. Which will cause the metal casing to bend a bit.


They could deal with the pressure by having a slight vacuum in the drive so that under the worst conditions, hot drive on a plane, the casing would be equal inside/outside pressure. Most constructs can be more easily built to withstand greater pressure on the outside rather than inside. Using good construction techniques tensile stresses can be translated to compressive stresses when you have greater pressure on the outside rather than inside of a container. Of course for the relatively flat top and bottom of the drive those force (and hence deflection) would be the same. But I have a feeling that the pressure difference we are talking about here, coupled with the strength of the container would show negligible deflection.
 
I do have secondary spindles. Heck, I even used to have a 1.5 TB Green drive and while I could hear it spin up after being idle, it was not so annoying that I'd want to eliminate 3.5" drives for 2.5" drives at this stage.

Indeed, but it is annoying enough that I disabled spindown in windows.
 
Hey, a correct statement from taltamir! Not something you see every day! :biggrin:

I think this behavior is what he was talking about.

I honestly wonder why mods don't ban you or give you infractions more often for your behavior.

If you check the date on the post, you'll see that it's 7 months old. This was dealt with months ago. Furthermore lashing out at another member in this fashion only makes the problem worse.
-ViRGE
 
Last edited by a moderator:
Back
Top