Not until SSD's offer the same or better capacity/$. SSD's don't make sense for archival data or large amounts of data that does not need to be accessed at higher speeds.
Well, HDDs barely make sense for archival data either, you still can get tapes at less than a fifth of HDD prices for the same capacity.
On the other hand, I think SSD builders do keep an ace in the hole by currently not offering 3.5" SSDs. As soon as the prices for NAND-Chips drop into the right region, they do have the chance to drop the $/GB ratio quickly by just adding a significantly higher number of NAND-Chips to a single controller and DDR-Cache.
HDDs won't vanish within the year, to be sure, but I would somewhat agree to Mr. Ogre in saying that the HDD will be running out of usefullness in the near future.
If we get multiple terabytes per square inch, I wonder if tapes can even keep up. Surely some of these advancements (like shingles) can be brought over to tape, too.
Whoops, didn't read lol. Still, even if SSDs are no problem for client applications, no level headed commercial application requiring long term data retention will choose solid state storage.
Sorry but BS, especially with respect to rebuilding arrays. And with 3DXPoint, solid state data retention will extend to 5x the length the very best magnetic platter is capable of.
saying a post BS while you saying BS ? 3DXPoint is still new, it have yet to prove it really give 5x better data retention... if already proved then share some legit data.
Although it not much cases that a drive need to be stored offline for longtime AFAIK there still none NAND on consumer grade that have data retention as long as magnetic platter.
With 2.5" SSD's available today offer lower power, higher capacity, higher performance and higher density than 3.5" HDD's. I wonder how much that offsets the higher cost per GB?
Higher capacity? A 10TB 2.5" SSD for <£500? Where?
Lower power than an HDD that's powered off?
Performance doesn't always matter.
The article says several times, this is not about desktop. This is about data-centre, extreme capacity, price-sensitive. These HDDs are competing with magnetic tape, not SSD.
You don't seem to understand how datacenters work. SSDs and modern JBOD infrastructure are changing the way this is approached. The thing you gotta realize is you can pack SSDs INSANELY dense. Yes, the power difference of 1 ssd is menial, but when you fill a rack with them, the combined power and cooling savings add up, especially if you are aiming for a minimum 2-3 year run cost.
You also have to bear in mind that datacenters are pretty much exclusively using substantially more expensive (and hotter/louder) SAS drives.
Capex is important, but you are totally ignoring Opex and TCO. Also AFR is about 1/6th (.5% vs 3%) which gives you more flexibity in your planing for consistency/redundancy. SSD failures are more or less predictable.
Either way the move to SSDs in the datacenter is VERY real, as density is king. Why waste money expanding into more datacenters and adding more racks? SSDs also solve alot of problems that HDD have such as large array rebuilds.
Also tech like RDMA combined with NVMe virtualization is going to fundamentally change the landscape.
First off, GREAT article Anton. This is what AT is all about.
I don't have a single HDD in my house anymore. Between 11TB on AWS and 800GB in OneDrive, it all comes down to the data centers which will all be using this technology.
Meanwhile the 480GB SSD's that cost $100 running my PC's and laptop have made magnetic storage irrelevant for my consumer use, so who can blame Seagate for not targeting me?
What everyone seems to be missing it NASes for the home with LARGE MEDIA collections. When you're looking at 2 GB for DVD rips and 4-5 GB for Blu Ray rips you need Terabytes of storage for $30/terabyte (or less hopefully) that SSDs can't touch. Even for full Blu Ray rips (some people want this) you're still looking at only needing 50 Mbps without any compression and even a lossless rip with Makemkv will take it down to 20 GB and will easily stream from a NAS with any decent spinning drive. When SSDs which are currently around $200/GB (for a consumer commercially available drive) to compete with spinning drives (say 3 TB @ $94 for a Toshiba or 3 TB for WD Red @ $109) then we won't see much in the consumer space. Not to even talk about 8 TB drives for around $200-$250. We are a long way from the demise of consumer spinning drives.
No kidding! I love my SSDs, but they are not going in my Nas any time soon. I have 5 3tb drives in a raid 6...that would cost a mint in SSDs still. Maybe I'll get there eventually, but it is going to be a long time.
Still, it is a sin to sell a pc with a hdd as a system drive these days. Really wish manufacturers would stop that
@JoeyJoJo123 >It's ogre. HDDs are dead. SSDs won. >Just give up.
Know how I know your data isn't backed up? There are two types of people. Those who have lost all their data, and those who are going to (possibly again).
Sure, you might not know how Amazon is storing your data, but I'll bet that it just isn't backed up.
Hard drives just moved up the data hierarchy. They are now stuck between tape (near the cost/TB without the huge entry costs) and SSD. They are also still ideal for NAS, especially consumer grade to fairly decent grade. Don't forget all the random access advantages HDDs have over tape.
I also expect to watch the whole HDD/SSD dance play over again with SDD/3dxpoint. Although the more interesting story will be if 3dxpoint will be able to replace main/virtual memory, leaving multi-GB DRAM "caches" between the CPU and "main memory".
Perhaps Mr. Ogre needs to consult with the tri-lambs. I'm sure they could find some uses for HDDs.
SMR was never a technology that ever interested me. I don't want more storage with measurably worse performance than existing ones. Seagate should have invested more into SSD tech than in SMR for HDDs.
You can do both. SMR for the platters with additional flash cache for frequently accessed data. Unfortunately HDDs are a one trick pony now, and they need to push in the one area they're better than SSDs in (relative amount of storage per $ spent).
Sure, you can try to mitigate the performance, but I feel the initial performance design should have been aiming higher than existing PMR performance. That's why SMR is disappointing to me. I feel like they were dumping R&D into HAMR, and they weren't getting results as fast as they wanted. So they go the SMR route to get the desired results (more storage density) in the interim.
It's not like SSDs. While SSDs were lower in capacity initially, the speed increase was dramatically better. SMR-based drives are only slightly better in capacity, but they are more noticeably worse in performance. Thus it didn't have the same impact - it doesn't have enough of the "wow" speed factor to help overlook its performance shortcomings.
It's like when Honda introduced the new Insight hybrid. It wasn't better or even at the same level of "MPG" as the Prius; that's why it didn't fare well.
Well, SMR is a stopgap solution, but it's here now and it works. HAMR sounds great on paper, but show me where I can buy an 8 TB HAMR drive today, for under $250?
My 8TB Seagate Archive has the lowest cost/GB of any drive at the time, and it works fine as a secondary storage/backup drive. Just manually dumping files to it or running scheduled backups work great, with performance that doesn't really "feel" any different than my 5900 RPM 4 TB PMR drive.
I also say this coming from someone who liked the Seagate 600 Pro SSD. They could have done a lot more years ago to compete for the market lead in consumer SSDs; basically Samsung dominates right now.
Even as an individuaI, I mostly don't care about storage performance. OS and apps and "live" data files get an SSD. That <5% of my storage. Media, archives and backups get an HDD. That's >95% of my storage. I don't care about performance for SSD nor HDD, both are OK for the uses I dedicate them too even in their cheapest crappiest incarnations. I'm more interested in cost, space, reliability, and durability.
Amen to that, beyond the tiny but vocal enthusiast community there are loads of users who want cheap, huge and reliable storage. I hope the SSD and HDD manufacturers don't forget this market and just concentrate on enthusiasts or enterprise customers.
I wonder how much of this will ever reach market? as SSD's take over ever more of the storage market the remaining HDD manufacturers will be required to spend ever more on R&D while dealing with shrinking revenue.
The consumer HDD market is slowly imploding, but while it's roughly half of all drives sold that's not where the money is. The Enterprise HDD market is doing better. Sales are down a bit; but while Seagate/WD/Toshiba don't break the two segments of enterprise drives out, it's probably from SSDs chewing into 10/15k drives in servers not the high capacity 3.5" drives used for bulk storage. Seagates comments about the next gen of 15k drives potentially being the last tends to back this up, since they're talking about multiple generations of other drive types.
Bulk data storage will stay with spinning rust until the price per GB crosses over in SSDs favor. Estimates I've seen on that a year or so ago were looking at 2025; but there's a fair amount of speculation there since the results of a half dozen generations of tech in each platform are somewhat speculative.
That patent [1] is clear example how current patent system is out of control and not useful to society at large. Short recap, patents original idea was to give temporal monopoly to inventor who disclosed his invention. There was no right to own anything you invented, but it was seen that disclosing the invention would ultimately benefit society who, after temporal monopoly, could use the invention.
Now opposite is true, there is nothing useful disclosed in Seagates patent! But we grand Seagate sole ownership to use Gold with Cu, Rh, Ru, ... or Mo in concentrations of 0.5-30% in HAMR NFT. How does this benefit us?
Of course Seagate can't stop WD to use same alloys as WD have similar patents to sue back, but they both can stifle smaller competitors. I don't think there will be any more competitors in HDD space, but same applies to other fields and their patents.
For consumer devices: noise. 10/15k HDDs are obnoxiously loud like an >40x CD or >16x DVD drive; except that they're spinning constantly not just for the few minutes it takes to read/write them. On the enterprise side, I'd guess implementation difficulty; probably due to vibrations was the limiting factor. If not mechanical strength of the platters themselves was probably the issue. Top end centrifuges can go to at least 70k RPM; but can be built much more heavily than a thin platter can.
Not just noise, but heat. In a constrained system like a laptop or desktop, the spindle and vibrations cause a ton of heat to be generated. Just look at WD's last consumer 10k drive, the 1 TB Velociraptor. It's a 2.5" drive that comes with its own heatsink for 3.5" bays. You definitely can't put that in a laptop, and it hits a market in the desktop space which has pretty much been taken over by SSDs.
I'm not sure about this. On my laptop, I've got a 120 GB SSD I got ages ago along with a recent 2 TB drive. Both cost about the same when new. I think HDDs will continue to get bigger and cheaper faster than SSDs, at least until some new process tech allows for very high flash densities and low production cost.
The main issue is price. There are still lots of users (like me) who can't afford huge multi-TB SSDs to hold everything, so they have to make do with a boot SSD and a storage HDD.
I'm honestly fed up with the poor reliability of HDDs. Of the three ones I still had in service, two are dead, and I'll never buy those two brands again.
This was in the back of my mind the whole time I read this article. What good is any of this if the damn drives keep dying? Maybe things are different on the enterprise side but I've given up on Seagate drives.
I like most have move to use SSD exclusively for the, hmm, "boot drive", But I do use HDD's for media storage where only the biggest highest density drive works. SSD's are currently about an order of magnitude more expensive per byte in this area. At a wild guess SSD's might reach price parity in a decade, but for now the HDD's are the only sane choice.
I'm currently struggling to fit in 6TB. My guess is 10Tb would be enough for the foreseeable future, and 20Tb would cover everything I am every likely to need. So I rather pleased to see Mr Re say they will hit 20Tb in a few years.
Samsung is hitting 15TB this year, so Seagate is done in enterprise. The performance/watt/$ is just vastly superior for SSDs. Once storage density is also in the wheelhouse of SSDs, the scales will tip and never go back.
This paragraph was enlightening - HDDs that use shingled recording write new tracks that overlap part of the previously written magnetic tracks. The overlapping tracks may slow down writing because the architecture requires HDDs to write the new data and then rewrite nearby tracks as well.
Which explains why my 5tb seagate with this tech can't seem to get past 40MB/s when writing to the drive.
Then read this - Ultimately, environments that involve a decent amount of writing might not be impressed with SMR performance, but the key figure here is density.
Is anyone getting a high capacity drive going to be impressed with 2-40 MB/s?
While all this new tech sounds great, all I want out of a HDD is reliability. In my house, HDDs are now just used for backup - I would gladly sacrifice performance, and even density, just to get a drive that I know will last 10 years. I still have G1 Intel SSDs from 2010 that are working fine as OS drives - so it's just a matter or GB/$ why I don't back up to SSD. It still seems like HDDs are nit-or-miss when it comes to reliability.
I have to agree to that first comment that they just have to give up. They are just slowing down their demise.
Currently and years to come, HDDs will become archival devices were fewer consumer devices will have HDDs in the coming years. The company will need this volume/scale from consumers to lessen the price of enterprise grade hardware.
If they cannot lessen the price for these enterprise customers, they will prefer SSDs despite costing more or having less capacity due to its higher performance and lower power consumption.
"While the evolution of consumer HDDs in the recent years was not fast, things are about to change."
The thing is, nobody needs huge consumer HDDs anymore. The only reason people ever bought large drives was for storing pictures/backups and for their pirated music/film collection. Well cheap/affordable digital streaming services have made a massive dent in piracy, and cloud storage/backup services provide excellent storage.
I honestly don't know any consumers who use more than 1-2TB of mechanical storage these days unless they are still hording all their VCD/DivX rips and .iso's...
You gotta wonder what's the BOM cost to produce a 0.5TB HDD vs Samsung's single chip 0.5TB SSD. And at 1TB, 2TB, etc. Are HDDs necessarily cheaper, being so much heavier and mechanically complex?
It's complicated. There's a minimum cost to build any mechanical hard drive, and the cheapest possible flash storage device is cheaper than that. That's why you don't see HDDs in any applications that need less than 128 GB or so anymore. It's just that the marginal cost of additional capacity is lower with hard drives (at least, until you start figuring in complex TCO arguments for data centers with very large amounts of storage).
It is almost irrelevant due to economies of scale. Right now, HDD are still comfortable with their scale right now. They have to be scared as it might drop non-linear in a few years; their drives will become more expensive that they will have to produce fewer, larger, archival drives.
Hard drives will remain as long as price/GB stays well ahead of solid state.
Seeing as manufacturers have been unwilling to maintain that for the last 4 years, all I can see is a death spiral.
I bought 3x 2TB HDDs in 2012 for $99NZD each. The current cheapest 2TB HDD is $115. And the cheapest HDD or GB is a 3TB at $156 (slightly more/GB).
Fark HDD manufacturers. They can't win on performance or reliability against Samsung SSDs, all they can do is win on price but that gap has shrunk so much in four years I don't hold much hope for them without a drastic move. SSDs don't have to reach the same price point to take over, they just need to be close enough when accounting for the performance gap.
It also doesn't help that they've already lost the density crown and in a much smaller form factor.
Just now (7/12/16 @ 19:17EST) heard on The Nightly Business Report (PBS) that Seagate: - announced earnings that exceeded analyst expectations - raised their projected income guidance - experienced their "best day ever" in the markets (measurement metrics not disclosed)
... oh yeah ...
and Seagate announced that they're eliminating 6,000 jobs ...
The future is certainly looking rosy ... I guess ...
I hear Seagate is looking to close out a major branch location nearing the end of the year or start of Q1 2017. Location in Shakopee, MN or a China branch (sorry, I didn't get what location in China.....rumor is pointing to MN location so far and as early as start of December has been hinted).
Makes you wonder how well things can really be going for them - unless this is a restructure to help them stay viable in the every fast changing world of storage technology.
So I was having a meeting with one of the large corporate storage providers asking when they thought the "mass change" would happen from HHD to SSD for corporate storage. Interestingly even their technology people were surprised, but at the enterprise level the inflection point happened this year for primary storage devices. Given much better TCO over time and vast performance improvement with SSDs they have better short and long term ROI due to a better than expected 40% decrease in price per TB in the past 12 months and the large jump in maximum enterprise SSD size. According to their people they are pushing all customers as fast as possible to SSD for all hot, and most warm storage. For them it makes perfect sense since they have very few drives break, arrays can easily be upgraded lowering the number of RFPs per $ sold and the performance increase for the customers is frankly amazing in the enterprise compute environment. I wonder how this shift will affect us on the bottom, since large storage providers suck a large portion of the overall storage market. They may be the 600 gorilla that forces the market to make a quick move off of HDDs, I would also be curious how the better performance and longer life span of SSDs will hurt SSD/HHD producers over the long haul since there will be much fewer sales over the long term.
what i want from HDD´s is REABILITY, REABILITY, REABILITY.
it can´t be that 10+% o harddrives from seagate die on me in the first 6-8 month. i have a small renderfarm at home and 8 sytems with 2-4 harddrives each.
infant mortality of HDD´s is a real problem. and overall REABILITY seems to be getting worse.
Make the hardware format size of a standard hard drive longer. There, you can now fit in two pairs of disks and double the archival storage. Where there is a will, there is a way. :)
We’ve updated our terms. By continuing to use the site and/or by logging into your account, you agree to the Site’s updated Terms of Use and Privacy Policy.
Facebook
Twitter
RSS
91 Comments
Back to Article
JoeyJoJo123 - Wednesday, July 6, 2016 - link
It's ogre. HDDs are dead. SSDs won.Just give up.
pancakes - Wednesday, July 6, 2016 - link
Not until SSD's offer the same or better capacity/$. SSD's don't make sense for archival data or large amounts of data that does not need to be accessed at higher speeds.ShieTar - Wednesday, July 6, 2016 - link
Well, HDDs barely make sense for archival data either, you still can get tapes at less than a fifth of HDD prices for the same capacity.On the other hand, I think SSD builders do keep an ace in the hole by currently not offering 3.5" SSDs. As soon as the prices for NAND-Chips drop into the right region, they do have the chance to drop the $/GB ratio quickly by just adding a significantly higher number of NAND-Chips to a single controller and DDR-Cache.
HDDs won't vanish within the year, to be sure, but I would somewhat agree to Mr. Ogre in saying that the HDD will be running out of usefullness in the near future.
mkozakewich - Thursday, July 7, 2016 - link
If we get multiple terabytes per square inch, I wonder if tapes can even keep up. Surely some of these advancements (like shingles) can be brought over to tape, too.cm2187 - Wednesday, July 6, 2016 - link
Plus I think you can kiss goodbye your data if you leave an SSD 2 years unpoweredMorawka - Wednesday, July 6, 2016 - link
SSD's under the endurance rating typically last over 10 years. There is even a article here on anandtech about it.tabascosauz - Wednesday, July 6, 2016 - link
SSD data retention != SSD endurance. There is even an article here on Ananadtech about it, go read it.tabascosauz - Wednesday, July 6, 2016 - link
Whoops, didn't read lol. Still, even if SSDs are no problem for client applications, no level headed commercial application requiring long term data retention will choose solid state storage.patrickjp93 - Wednesday, July 6, 2016 - link
Sorry but BS, especially with respect to rebuilding arrays. And with 3DXPoint, solid state data retention will extend to 5x the length the very best magnetic platter is capable of.slyphnier - Friday, July 8, 2016 - link
saying a post BS while you saying BS ?3DXPoint is still new, it have yet to prove it really give 5x better data retention... if already proved then share some legit data.
Although it not much cases that a drive need to be stored offline for longtime AFAIK there still none NAND on consumer grade that have data retention as long as magnetic platter.
mkozakewich - Thursday, July 7, 2016 - link
Oh, I remember that article. Higher write temperatures mean better longevity, right?twelvebore - Wednesday, July 6, 2016 - link
Guessing that you don't buy storage by the petabyte then? You know, horizons and all that.Ushio01 - Wednesday, July 6, 2016 - link
With 2.5" SSD's available today offer lower power, higher capacity, higher performance and higher density than 3.5" HDD's. I wonder how much that offsets the higher cost per GB?twelvebore - Wednesday, July 6, 2016 - link
Higher capacity? A 10TB 2.5" SSD for <£500? Where?Lower power than an HDD that's powered off?
Performance doesn't always matter.
The article says several times, this is not about desktop. This is about data-centre, extreme capacity, price-sensitive. These HDDs are competing with magnetic tape, not SSD.
jwhannell - Wednesday, July 6, 2016 - link
Flape.patrickjp93 - Wednesday, July 6, 2016 - link
Performance/Watt/$ is the most important metric, and HDD is already under immense pressure from archival SSDs.patrickjp93 - Wednesday, July 6, 2016 - link
For enterprise use that's a $2000 drive, unless you're using one without power loss protection and ECC... And Samsung already has one provided.Murloc - Wednesday, July 6, 2016 - link
It doesn't offset the cost at all if the only thing that matters is $/GB.amnesia0287 - Wednesday, July 6, 2016 - link
You don't seem to understand how datacenters work. SSDs and modern JBOD infrastructure are changing the way this is approached. The thing you gotta realize is you can pack SSDs INSANELY dense. Yes, the power difference of 1 ssd is menial, but when you fill a rack with them, the combined power and cooling savings add up, especially if you are aiming for a minimum 2-3 year run cost.You also have to bear in mind that datacenters are pretty much exclusively using substantially more expensive (and hotter/louder) SAS drives.
Capex is important, but you are totally ignoring Opex and TCO. Also AFR is about 1/6th (.5% vs 3%) which gives you more flexibity in your planing for consistency/redundancy. SSD failures are more or less predictable.
Either way the move to SSDs in the datacenter is VERY real, as density is king. Why waste money expanding into more datacenters and adding more racks? SSDs also solve alot of problems that HDD have such as large array rebuilds.
Also tech like RDMA combined with NVMe virtualization is going to fundamentally change the landscape.
zodiacfml - Friday, July 8, 2016 - link
Correct. SSDs have higher density already and the rich companies can afford them.Zak - Wednesday, July 6, 2016 - link
Forget enterprise. I use 4GB drives as local backups and planning to go up to 6 or 8. Show me affordable 8TB SSD I can use for backup.inighthawki - Wednesday, July 6, 2016 - link
8TB no, but I'm sure I can find you a few good 4GB drives :)cm2187 - Wednesday, July 6, 2016 - link
WD Reds?cm2187 - Wednesday, July 6, 2016 - link
Actually even cheaper if it is for backups: seagate 8tb archive drives.Samus - Wednesday, July 6, 2016 - link
First off, GREAT article Anton. This is what AT is all about.I don't have a single HDD in my house anymore. Between 11TB on AWS and 800GB in OneDrive, it all comes down to the data centers which will all be using this technology.
Meanwhile the 480GB SSD's that cost $100 running my PC's and laptop have made magnetic storage irrelevant for my consumer use, so who can blame Seagate for not targeting me?
trivor - Thursday, July 7, 2016 - link
What everyone seems to be missing it NASes for the home with LARGE MEDIA collections. When you're looking at 2 GB for DVD rips and 4-5 GB for Blu Ray rips you need Terabytes of storage for $30/terabyte (or less hopefully) that SSDs can't touch. Even for full Blu Ray rips (some people want this) you're still looking at only needing 50 Mbps without any compression and even a lossless rip with Makemkv will take it down to 20 GB and will easily stream from a NAS with any decent spinning drive. When SSDs which are currently around $200/GB (for a consumer commercially available drive) to compete with spinning drives (say 3 TB @ $94 for a Toshiba or 3 TB for WD Red @ $109) then we won't see much in the consumer space. Not to even talk about 8 TB drives for around $200-$250. We are a long way from the demise of consumer spinning drives.CaedenV - Thursday, July 7, 2016 - link
No kidding! I love my SSDs, but they are not going in my Nas any time soon. I have 5 3tb drives in a raid 6...that would cost a mint in SSDs still. Maybe I'll get there eventually, but it is going to be a long time.Still, it is a sin to sell a pc with a hdd as a system drive these days. Really wish manufacturers would stop that
JlHADJOE - Wednesday, July 6, 2016 - link
IBM showed us that magnetic storage can store a bit using as few as 12 atoms. That's far denser than any type of memory developed so far.http://www.wired.com/2012/01/ibm-scientists/
SSDs will replace HDDs for most of the consumer market, but HDDs will stay around for bulk data.
Cygni - Wednesday, July 6, 2016 - link
It's 'ogre'? Is shrek around or something?Also if you read the article, you will see that this isn't exactly focused at the same market as enthusiast SSDs.
Michael Bay - Wednesday, July 6, 2016 - link
Tell that to 8Tb of media I have copying to the new HDD now.abrowne1993 - Wednesday, July 6, 2016 - link
Is that all for one second of CGI in the new Transformers movie?Michael Bay - Thursday, July 7, 2016 - link
And some porn.Holliday75 - Friday, July 8, 2016 - link
Is Megan Fox back? She could probably use the cash.JlHADJOE - Monday, July 11, 2016 - link
8 Terra bits? So 1TB =)wumpus - Thursday, July 7, 2016 - link
@JoeyJoJo123>It's ogre. HDDs are dead. SSDs won.
>Just give up.
Know how I know your data isn't backed up? There are two types of people. Those who have lost all their data, and those who are going to (possibly again).
Sure, you might not know how Amazon is storing your data, but I'll bet that it just isn't backed up.
Hard drives just moved up the data hierarchy. They are now stuck between tape (near the cost/TB without the huge entry costs) and SSD. They are also still ideal for NAS, especially consumer grade to fairly decent grade. Don't forget all the random access advantages HDDs have over tape.
I also expect to watch the whole HDD/SSD dance play over again with SDD/3dxpoint. Although the more interesting story will be if 3dxpoint will be able to replace main/virtual memory, leaving multi-GB DRAM "caches" between the CPU and "main memory".
Perhaps Mr. Ogre needs to consult with the tri-lambs. I'm sure they could find some uses for HDDs.
Nagorak - Monday, July 11, 2016 - link
I'm having difficulty understanding your comment but if you're implying that data stored on AWS isn't backed up then I find that almost inconceivable.hectorsm - Thursday, July 7, 2016 - link
You're delusional. Today ~99% of PC shipments still have HDDs.romrunning - Wednesday, July 6, 2016 - link
SMR was never a technology that ever interested me. I don't want more storage with measurably worse performance than existing ones. Seagate should have invested more into SSD tech than in SMR for HDDs.JoeyJoJo123 - Wednesday, July 6, 2016 - link
You can do both. SMR for the platters with additional flash cache for frequently accessed data. Unfortunately HDDs are a one trick pony now, and they need to push in the one area they're better than SSDs in (relative amount of storage per $ spent).extide - Thursday, July 7, 2016 - link
Yeah I always thought the combo of a SMR drive + some NAND cache was like super obvious, and I am wondering why we haven't seen more drives like that.nandnandnand - Wednesday, July 6, 2016 - link
They explained how to mitigate SMR performance issues in the article.romrunning - Wednesday, July 6, 2016 - link
Sure, you can try to mitigate the performance, but I feel the initial performance design should have been aiming higher than existing PMR performance. That's why SMR is disappointing to me. I feel like they were dumping R&D into HAMR, and they weren't getting results as fast as they wanted. So they go the SMR route to get the desired results (more storage density) in the interim.It's not like SSDs. While SSDs were lower in capacity initially, the speed increase was dramatically better. SMR-based drives are only slightly better in capacity, but they are more noticeably worse in performance. Thus it didn't have the same impact - it doesn't have enough of the "wow" speed factor to help overlook its performance shortcomings.
It's like when Honda introduced the new Insight hybrid. It wasn't better or even at the same level of "MPG" as the Prius; that's why it didn't fare well.
JimmiG - Thursday, July 7, 2016 - link
Well, SMR is a stopgap solution, but it's here now and it works. HAMR sounds great on paper, but show me where I can buy an 8 TB HAMR drive today, for under $250?My 8TB Seagate Archive has the lowest cost/GB of any drive at the time, and it works fine as a secondary storage/backup drive. Just manually dumping files to it or running scheduled backups work great, with performance that doesn't really "feel" any different than my 5900 RPM 4 TB PMR drive.
romrunning - Wednesday, July 6, 2016 - link
I also say this coming from someone who liked the Seagate 600 Pro SSD. They could have done a lot more years ago to compete for the market lead in consumer SSDs; basically Samsung dominates right now.StormyParis - Wednesday, July 6, 2016 - link
Even as an individuaI, I mostly don't care about storage performance. OS and apps and "live" data files get an SSD. That <5% of my storage. Media, archives and backups get an HDD. That's >95% of my storage. I don't care about performance for SSD nor HDD, both are OK for the uses I dedicate them too even in their cheapest crappiest incarnations.I'm more interested in cost, space, reliability, and durability.
serendip - Friday, July 8, 2016 - link
Amen to that, beyond the tiny but vocal enthusiast community there are loads of users who want cheap, huge and reliable storage. I hope the SSD and HDD manufacturers don't forget this market and just concentrate on enthusiasts or enterprise customers.paulemannsen - Saturday, July 9, 2016 - link
+1. I could even live with 10/10 mb read/write speeds, just give me more reliability and space.Nozuka - Wednesday, July 6, 2016 - link
There are already 13TB SSDs... currently extremely expensive, but it's only a matter of time...XZerg - Wednesday, July 6, 2016 - link
you mean 16TB - from Samsung - http://gizmodo.com/samsungs-16tb-ssd-is-now-an-act...Lolimaster - Friday, July 8, 2016 - link
The 13TB was from fixstars (japan based). They were the 1st in multiterabyte SSDs 3-6-10-13TB.Ushio01 - Wednesday, July 6, 2016 - link
I wonder how much of this will ever reach market? as SSD's take over ever more of the storage market the remaining HDD manufacturers will be required to spend ever more on R&D while dealing with shrinking revenue.DanNeely - Wednesday, July 6, 2016 - link
The consumer HDD market is slowly imploding, but while it's roughly half of all drives sold that's not where the money is. The Enterprise HDD market is doing better. Sales are down a bit; but while Seagate/WD/Toshiba don't break the two segments of enterprise drives out, it's probably from SSDs chewing into 10/15k drives in servers not the high capacity 3.5" drives used for bulk storage. Seagates comments about the next gen of 15k drives potentially being the last tends to back this up, since they're talking about multiple generations of other drive types.Bulk data storage will stay with spinning rust until the price per GB crosses over in SSDs favor. Estimates I've seen on that a year or so ago were looking at 2025; but there's a fair amount of speculation there since the results of a half dozen generations of tech in each platform are somewhat speculative.
Anato - Wednesday, July 6, 2016 - link
That patent [1] is clear example how current patent system is out of control and not useful to society at large. Short recap, patents original idea was to give temporal monopoly to inventor who disclosed his invention. There was no right to own anything you invented, but it was seen that disclosing the invention would ultimately benefit society who, after temporal monopoly, could use the invention.Now opposite is true, there is nothing useful disclosed in Seagates patent! But we grand Seagate sole ownership to use Gold with Cu, Rh, Ru, ... or Mo in concentrations of 0.5-30% in HAMR NFT. How does this benefit us?
Of course Seagate can't stop WD to use same alloys as WD have similar patents to sue back, but they both can stifle smaller competitors. I don't think there will be any more competitors in HDD space, but same applies to other fields and their patents.
[1] http://www.google.com/patents/US8427925
[Disclosure, I didn't read full patent application, did cursory review and didn't find anything of value]
Zak - Wednesday, July 6, 2016 - link
I was wondering why HDs never broke the 7500rpm and 15000rpm barriers.DanNeely - Wednesday, July 6, 2016 - link
For consumer devices: noise. 10/15k HDDs are obnoxiously loud like an >40x CD or >16x DVD drive; except that they're spinning constantly not just for the few minutes it takes to read/write them. On the enterprise side, I'd guess implementation difficulty; probably due to vibrations was the limiting factor. If not mechanical strength of the platters themselves was probably the issue. Top end centrifuges can go to at least 70k RPM; but can be built much more heavily than a thin platter can.metayoshi - Wednesday, July 6, 2016 - link
Not just noise, but heat. In a constrained system like a laptop or desktop, the spindle and vibrations cause a ton of heat to be generated. Just look at WD's last consumer 10k drive, the 1 TB Velociraptor. It's a 2.5" drive that comes with its own heatsink for 3.5" bays. You definitely can't put that in a laptop, and it hits a market in the desktop space which has pretty much been taken over by SSDs.piasabird - Wednesday, July 6, 2016 - link
Still using hard drives. No reason to switch when hard drives are so cheap. When I watch TV on the Internet, It still works so no reason to change.pavag - Wednesday, July 6, 2016 - link
Kurzweil curse.Nobody accounts the relative speed of price reduction. SSD will get cheaper per gigabyte than SSD before 2020.
We will never see HDMR HDD. And there is a chance we will not even get HAMR.
And article speculating about the date when SSD will beat HDD for price per gigabyte would be interesting.
serendip - Friday, July 8, 2016 - link
I'm not sure about this. On my laptop, I've got a 120 GB SSD I got ages ago along with a recent 2 TB drive. Both cost about the same when new. I think HDDs will continue to get bigger and cheaper faster than SSDs, at least until some new process tech allows for very high flash densities and low production cost.The main issue is price. There are still lots of users (like me) who can't afford huge multi-TB SSDs to hold everything, so they have to make do with a boot SSD and a storage HDD.
jabber - Wednesday, July 6, 2016 - link
So you didn't bother to ask why Seagate's HDD reliability has nosedived since the floods a few years back?Notmyusualid - Wednesday, July 6, 2016 - link
Just lost my Seagate 4TB 2.5" internal disk last week.Luckily my most precious things were backed-up elsewhere, but dam, 2.4TB lost.
jwcalla - Thursday, July 7, 2016 - link
I'm honestly fed up with the poor reliability of HDDs. Of the three ones I still had in service, two are dead, and I'll never buy those two brands again.Michael Bay - Thursday, July 7, 2016 - link
Out of not so idle interest, what were they?kamm2 - Thursday, July 7, 2016 - link
This was in the back of my mind the whole time I read this article. What good is any of this if the damn drives keep dying? Maybe things are different on the enterprise side but I've given up on Seagate drives.jbrizz - Wednesday, July 6, 2016 - link
Can we just have 5.25 inch hard drives again? I don't care so much about density at home, but I need to hoard more files!rstuart - Wednesday, July 6, 2016 - link
I like most have move to use SSD exclusively for the, hmm, "boot drive", But I do use HDD's for media storage where only the biggest highest density drive works. SSD's are currently about an order of magnitude more expensive per byte in this area. At a wild guess SSD's might reach price parity in a decade, but for now the HDD's are the only sane choice.I'm currently struggling to fit in 6TB. My guess is 10Tb would be enough for the foreseeable future, and 20Tb would cover everything I am every likely to need. So I rather pleased to see Mr Re say they will hit 20Tb in a few years.
patrickjp93 - Wednesday, July 6, 2016 - link
Samsung is hitting 15TB this year, so Seagate is done in enterprise. The performance/watt/$ is just vastly superior for SSDs. Once storage density is also in the wheelhouse of SSDs, the scales will tip and never go back.gospadin - Thursday, July 7, 2016 - link
absolute density (TB/m3) is already in SSD's advantageThe biggest 2.5" HDD you can buy today is 2TB. In that form factor, Samsung is currently selling 4TB SSD, with 8TB/16TB drives announced.
Lolimaster - Friday, July 8, 2016 - link
Do like me:4x6TB WD blues, so should be covers for some time.
anactoraaron - Thursday, July 7, 2016 - link
This paragraph was enlightening -HDDs that use shingled recording write new tracks that overlap part of the previously written magnetic tracks. The overlapping tracks may slow down writing because the architecture requires HDDs to write the new data and then rewrite nearby tracks as well.
Which explains why my 5tb seagate with this tech can't seem to get past 40MB/s when writing to the drive.
Then read this - Ultimately, environments that involve a decent amount of writing might not be impressed with SMR performance, but the key figure here is density.
Is anyone getting a high capacity drive going to be impressed with 2-40 MB/s?
No. No they will not.
anactoraaron - Thursday, July 7, 2016 - link
Forgot to mention that it is a 5tb archive drive. Didn't realize that it would take me 3 days to archive 3tb to it.kmmatney - Thursday, July 7, 2016 - link
While all this new tech sounds great, all I want out of a HDD is reliability. In my house, HDDs are now just used for backup - I would gladly sacrifice performance, and even density, just to get a drive that I know will last 10 years. I still have G1 Intel SSDs from 2010 that are working fine as OS drives - so it's just a matter or GB/$ why I don't back up to SSD. It still seems like HDDs are nit-or-miss when it comes to reliability.Lolimaster - Friday, July 8, 2016 - link
4x6TB WD blue and get more drives to mirror data + maybe google drive 10TB plan?AnnonymousCoward - Friday, July 8, 2016 - link
Hitachi has significantly better reliability!http://www.extremetech.com/computing/228497-backbl...
Lolimaster - Friday, July 8, 2016 - link
Right now I wont touch a Seagate even with a 1000km pole. SMR is pretty much cr*ap for consumers. Where are my HAMR? 15-20TB drives?Michael Bay - Friday, July 8, 2016 - link
In 2020 somewhere.zodiacfml - Friday, July 8, 2016 - link
I have to agree to that first comment that they just have to give up. They are just slowing down their demise.Currently and years to come, HDDs will become archival devices were fewer consumer devices will have HDDs in the coming years. The company will need this volume/scale from consumers to lessen the price of enterprise grade hardware.
If they cannot lessen the price for these enterprise customers, they will prefer SSDs despite costing more or having less capacity due to its higher performance and lower power consumption.
James_Edge - Friday, July 8, 2016 - link
"While the evolution of consumer HDDs in the recent years was not fast, things are about to change."The thing is, nobody needs huge consumer HDDs anymore. The only reason people ever bought large drives was for storing pictures/backups and for their pirated music/film collection. Well cheap/affordable digital streaming services have made a massive dent in piracy, and cloud storage/backup services provide excellent storage.
I honestly don't know any consumers who use more than 1-2TB of mechanical storage these days unless they are still hording all their VCD/DivX rips and .iso's...
jabber - Friday, July 8, 2016 - link
Yeah I see the folks who have TBs of ripped anime/porn etc. etc. as having a mental illness. New form of packrat syndrome. Let it go.AnnonymousCoward - Friday, July 8, 2016 - link
You gotta wonder what's the BOM cost to produce a 0.5TB HDD vs Samsung's single chip 0.5TB SSD. And at 1TB, 2TB, etc. Are HDDs necessarily cheaper, being so much heavier and mechanically complex?drothgery - Saturday, July 9, 2016 - link
It's complicated. There's a minimum cost to build any mechanical hard drive, and the cheapest possible flash storage device is cheaper than that. That's why you don't see HDDs in any applications that need less than 128 GB or so anymore. It's just that the marginal cost of additional capacity is lower with hard drives (at least, until you start figuring in complex TCO arguments for data centers with very large amounts of storage).zodiacfml - Sunday, July 10, 2016 - link
It is almost irrelevant due to economies of scale. Right now, HDD are still comfortable with their scale right now. They have to be scared as it might drop non-linear in a few years; their drives will become more expensive that they will have to produce fewer, larger, archival drives.AnnonymousCoward - Sunday, July 10, 2016 - link
"almost irrelevant"? HDD's entire play is cost/GB, and there's reason to believe they cost MORE below some capacity."economy of scale" - you mean volume? Samsung already ships >10B memory chips per year, far more than Seagate units.
Danvelopment - Sunday, July 10, 2016 - link
Hard drives will remain as long as price/GB stays well ahead of solid state.Seeing as manufacturers have been unwilling to maintain that for the last 4 years, all I can see is a death spiral.
I bought 3x 2TB HDDs in 2012 for $99NZD each. The current cheapest 2TB HDD is $115. And the cheapest HDD or GB is a 3TB at $156 (slightly more/GB).
Fark HDD manufacturers. They can't win on performance or reliability against Samsung SSDs, all they can do is win on price but that gap has shrunk so much in four years I don't hold much hope for them without a drastic move. SSDs don't have to reach the same price point to take over, they just need to be close enough when accounting for the performance gap.
It also doesn't help that they've already lost the density crown and in a much smaller form factor.
Bobs_Your_Uncle - Tuesday, July 12, 2016 - link
Just now (7/12/16 @ 19:17EST) heard on The Nightly Business Report (PBS) that Seagate:- announced earnings that exceeded analyst expectations
- raised their projected income guidance
- experienced their "best day ever" in the markets (measurement metrics not disclosed)
... oh yeah ...
and Seagate announced that they're eliminating 6,000 jobs ...
The future is certainly looking rosy ... I guess ...
neatfeatguy - Thursday, July 14, 2016 - link
I hear Seagate is looking to close out a major branch location nearing the end of the year or start of Q1 2017. Location in Shakopee, MN or a China branch (sorry, I didn't get what location in China.....rumor is pointing to MN location so far and as early as start of December has been hinted).Makes you wonder how well things can really be going for them - unless this is a restructure to help them stay viable in the every fast changing world of storage technology.
truemore - Thursday, July 28, 2016 - link
So I was having a meeting with one of the large corporate storage providers asking when they thought the "mass change" would happen from HHD to SSD for corporate storage.Interestingly even their technology people were surprised, but at the enterprise level the inflection point happened this year for primary storage devices. Given much better TCO over time and vast performance improvement with SSDs they have better short and long term ROI due to a better than expected 40% decrease in price per TB in the past 12 months and the large jump in maximum enterprise SSD size.
According to their people they are pushing all customers as fast as possible to SSD for all hot, and most warm storage. For them it makes perfect sense since they have very few drives break, arrays can easily be upgraded lowering the number of RFPs per $ sold and the performance increase for the customers is frankly amazing in the enterprise compute environment.
I wonder how this shift will affect us on the bottom, since large storage providers suck a large portion of the overall storage market. They may be the 600 gorilla that forces the market to make a quick move off of HDDs,
I would also be curious how the better performance and longer life span of SSDs will hurt SSD/HHD producers over the long haul since there will be much fewer sales over the long term.
Gothmoth - Saturday, July 30, 2016 - link
what i want from HDD´s is REABILITY, REABILITY, REABILITY.it can´t be that 10+% o harddrives from seagate die on me in the first 6-8 month.
i have a small renderfarm at home and 8 sytems with 2-4 harddrives each.
infant mortality of HDD´s is a real problem. and overall REABILITY seems to be getting worse.
Gothmoth - Saturday, July 30, 2016 - link
sorry i meant RELIABILITY... :-)profquatermass - Monday, August 29, 2016 - link
Oh the Irony......profquatermass - Monday, August 29, 2016 - link
Here's an idea.Make the hardware format size of a standard hard drive longer. There, you can now fit in two pairs of disks and double the archival storage. Where there is a will, there is a way. :)