Western Digital is now shipping 14 TB hard drives. The products use shingled magnetic recording (SMR), which can slow down re-writes:
Western Digital has started to ship its new HGST Ultrastar Hs14 hard drives, promoted as being suitable for cloud datacenters and for hyperscale developments. The capacity increase from its predecessor, the Ultrastar Ha10, from 10TB to 14 TB offers a significant performance improvement. The new 14 TB HDD is based on shingled magnetic recording technology, which is a system that naturally focuses more on sequential write performance. These drives will only be available with host management, which means it will not be available to general consumers, but only to select customers of HGST.
The HGST Ultrastar Hs14 relies on Western Digital's fourth-generation HelioSeal enterprise platform which integrates eight platters and features various internal components specially designed for such hard drives. The new helium-filled HDD has a 7200 RPM spindle speed, a 512 MB cache. and numerous enhancements when it comes to reliability and durability of the drive. As with other HGST enterprise-class HDDs, the Ultrastar Hs14 is rated for 2.5 million hours MTBF and comes with a five-year warranty.
Previously: Western Digital Announces 12-14 TB Hard Drives and an 8 TB SSD
Seagate's 12 TB HDDs Are in Use, and 16 TB is Planned for 2018
Western Digital Begins Shipping 12 TB Helium-Filled Drives with 8 Platters
Seagate Launches Consumer-Oriented 12 TB Drives
Western Digital to Use Microwave Assisted Magnetic Recording to Produce 40 TB HDDs by 2025
(Score: 4, Funny) by DannyB on Thursday November 02 2017, @09:21PM (8 children)
Why not 10,000 RPM speed instead of 7,200 RPM?
With helium being such a tiny atom, how can they effectively seal the drive for long duration without the gas escaping? If the gas escapes does drive performance degrade or suffer in some weigh?
Wow 14 TB, that ought to be enough for anybody!
If I record my voice in an mp3 file and play it back will it sound funny due to being helium filled?
Young people won't believe you if you say you used to get Netflix by US Postal Mail.
(Score: 2) by fishybell on Thursday November 02 2017, @09:26PM (2 children)
Whoa, whoa there. Calm down. Take a breath. Step away from the keyboard. Take a nice calming walk outside, then come back and re-assess.
(Score: 2) by LoRdTAW on Thursday November 02 2017, @09:32PM
It's why I try to avoid posting high.
(Score: 2) by bob_super on Thursday November 02 2017, @09:34PM
Alternatively, consume all 14TB of porn therein, then come back and re-assess.
(Score: 0) by Anonymous Coward on Thursday November 02 2017, @09:29PM (1 child)
Price?
(Score: 2) by NotSanguine on Thursday November 02 2017, @09:45PM
That question is answered in TFS:
Translation: A lot.
No, no, you're not thinking; you're just being logical. --Niels Bohr
(Score: 1, Informative) by Anonymous Coward on Thursday November 02 2017, @09:39PM (1 child)
Probably reliability although you would think the helium could counteract. Also 10K makes much less sense because SSD has taken over speedy applications and this is for bulk, mostly cold storage.
Economics of high RPM HDDs discussed here and it ain't a pretty picture for the 10-15K RPM spinning rust:
https://www.theregister.co.uk/2017/08/10/128tb_ssds_signal_coming_armageddon_for_disk_drives/ [theregister.co.uk]
https://www.theregister.co.uk/2015/03/13/about_the_real_price_of_flash_and_disk/ [theregister.co.uk]
I would have to look it up but I think it still works after seal failure. Both WD and Seagate spent years developing the technology. Warranties are like 5 years so they must be confident in it.
No your voice will not sound funny and how dare you make a joke out of this srs bznss.
~t
(Score: 3, Funny) by DannyB on Friday November 03 2017, @01:46PM
I would never make a joke on Soylent.
Young people won't believe you if you say you used to get Netflix by US Postal Mail.
(Score: 0) by Anonymous Coward on Friday November 03 2017, @02:24PM
Former employee at an HDD company:
10,000 RPM burns too much power, customers prefer slower to more power hungry.
Helium is kept in with laser welds. Yes, I'm unsure as Helium wasn't my thing, but it could actually be a catastrophic failure, hence the laser welds.
I heard that same thing said of 1GB long ago.
No, no it will not.
(Score: 2) by bzipitidoo on Thursday November 02 2017, @10:49PM (3 children)
How long do these drives last? Lately, I have had too many HDDs die young on me. I'd rather have a 4TB drive that lasts 10 years (if there is such a thing) than this if it's going to fail in 3 years. The "TB years" is nearly equivalent: 4*10 = 40 vs 14*3 = 42.
(Score: 0) by Anonymous Coward on Thursday November 02 2017, @10:53PM (1 child)
"the Ultrastar Hs14 is rated for 2.5 million hours MTBF and comes with a five-year warranty."
(Score: 1) by toddestan on Sunday November 05 2017, @06:32AM
The MTBF number is marketing bullshit, the actual MTBF is likely well below that. However, should it fail within 5 years I would trust them to honor the warranty.
(Score: 2) by forkazoo on Friday November 03 2017, @12:39AM
Most folks using something like this will have a very large number of spares on hand to swap into their RAID arrays or zPools or whatever. It won't really be targeted to people who are going to buy one and notice if it fails.
(Score: 2) by requerdanos on Thursday November 02 2017, @11:29PM (4 children)
Given that SMR records one thing partially underneath another underneath another, and all those others have to be moved out of the way to re-write something old, saying that it "can slow down" rewrites is sort of like saying that filling your car with lead "can potentially result in lower top speeds and possibly lower gas mileage."
(Score: 2) by KiloByte on Friday November 03 2017, @12:41AM (1 child)
Also, SMR really, really sucks when used with a regular filesystem. Without specific tuning for SMR, f2fs fares best by a good margin despite having been designed for a completely different use case (flash with weak FTL). In theory, btrfs should shine as well but this is not the case -- and no one implemented such SMR tuning yet.
You want random writes (even to structs like inodes) be converted into a long linear write, appending only within a specific zone.
Ceterum censeo systemd esse delendam.
(Score: 2) by takyon on Friday November 03 2017, @12:47AM
This is an enterprise bulk storage product. Although WD implies that this could be used to replace non-shingled 10 TB drives.
[SIG] 10/28/2017: Soylent Upgrade v14 [soylentnews.org]
(Score: 2) by takyon on Friday November 03 2017, @12:52AM (1 child)
I strongly hedge when writing summaries. "which can slow down re-writes" replaced something like "which has lower write speeds" in draft 1. In fact, initial sequential writes for this product are up by a lot from the Ultrastar Ha10.
[SIG] 10/28/2017: Soylent Upgrade v14 [soylentnews.org]
(Score: 2) by requerdanos on Friday November 03 2017, @11:27AM
Ah, my post was part clarification, part silliness. No criticism intended--and you're right, rewrites are the case where SMR doesn't shine (not writing in general).