bit-tech.net

Hard drive market set for double-digit slump

Hard drive market set for double-digit slump

The market for traditional mechanical drives, while massive, is shrinking, with one market watcher predicting a near-12 per cent drop in revenue for 2013.

The global market for traditional spinning-platter hard disk drives is set to take a massive slump this year, with industry analysts pointing to a double-digit drop in revenue.

According to a report from industry watcher IHS iSuppli, the hard drive market's revenue for the year will drop 11.8 per cent to a still pretty sizeable $32.7 billion, compared to $37.1 billion in the last financial year. Things won't get much better the year after, with iSuppli predicting a slight slip to $32 billion for 2014.

The reasons for such a gloomy outlook? The growing domination of solid-state storage devices and the increasing interest in so-called 'post-PC' devices like tablets and smartphones. 'The HDD industry will face myriad challenges in 2013,'' claimed Fang Zhang, analyst for storage systems at IHS, of the report. 'Shipments for desktop PCs will slip this year, while notebook sales are under pressure as consumers continue to favour smartphones and tablets. The declining price of SSDs also will allow them to take away some share from conventional HDDs.'

It's worth pointing out that iSuppli isn't predicting the death of the mechanical hard drive, however: the company's report points out that spinning platters are still king of the hill when it comes to cost per gigabyte, and it's increasingly common to see original equipment manufacturers (OEMs) and original design manufacturers (ODMs) exploit that fact by pairing a spinning disk with an SSD or using hybrid drives, even in ultra-slim devices like the current crop of Intel Ultrabooks.

While sales may be slipping at the cheaper end of the consumer market, where buyers are increasingly turning to tablets for their computing needs rather than upgrading their ageing laptops or desktops, iSuppli's report claims that enterprise use - in particular for cloud storage systems and 'big data' server farms - is only set to grow, being a market where solid-state disks would be cost prohibitive for the capacity required.

As a final prediction that should surprise no-one, iSuppli points to a potential end to the optical drive. As the world increasingly looks to the internet for distribution of software, video and audio, the optical drive - and, as HMV has discovered to its cost, the weekly journey to the shops to check out the new releases - is looking under increasing threat. As new ultra-compact form factors - like the aforementioned Ultrabook and Intel's Next Unit of Computing (NUC) desktop design - push the drive out as an external component, it's likely that fewer consumers will feel the pinch of its absence.

For consumers, the news of a slowing hard drive market is a mixed bag: iSuppli predicts prices could slip by as much as seven per cent over the coming year, meaning cheaper storage for all - but tighter revenues will lead to reduced research and development spending at the major manufacturers along with 'inventory adjustments' that may see production slowed to keep prices artificially high.

43 Comments

Discuss in the forums Reply
Mankz 5th February 2013, 10:15 Quote
The prices haven't fallen since the floods back to their original levels, while the prices and speeds of SSD's continue to fall to keep making them more attractive options.. that and more cloud based storage.
Gareth Halfacree 5th February 2013, 10:35 Quote
Quote:
Originally Posted by Mankz
[...] that and more cloud based storage.
And onto what does the cloud store that data? :p
sniperdude 5th February 2013, 10:40 Quote
I want drives but I will wait till the prices drop
Dudey109 5th February 2013, 10:46 Quote
Quote:
Originally Posted by Gareth Halfacree
And onto what does the cloud store that data? :p

fluffy white clouds above our head. no HDDs required.
runadumb 5th February 2013, 10:48 Quote
I'm after a 4TB drive for my NAS and maybe a new 2TB Desktop drive to replace my aging (and nearly full) 1TB one. Although as its mostly full of Steam games getting a 512GB SSD could also be an option...but maybe next year.
jrs77 5th February 2013, 11:56 Quote
Couldn't care any less, as I don't have tons of data to begin with.

The workfiles from the last 10 years take up some 20GB (PSD, HTML, AI, JPG, etc), my MP3-files use some 60GB, and I'm not collecting much movies, maybe some 4-500GB of series and movies.

A 1 TB NAS (backup) + 32 GB (workfiles) and 64 GB (music) USB-stick is all I need really to store and backup my files :p
theshadow2001 5th February 2013, 12:05 Quote
Quote:
Originally Posted by Gareth Halfacree
And onto what does the cloud store that data? :p

True. But, where as before 5 users would have 5 hard drives, with cloud storage 5 users could share a single cloud hard drive. Its a more efficient use of a single disk.

As for R&D on hard drives, I say screw it. Lets just move to SSDs make em big and cheap. Hard drives will eventually go the way of the dodo.
Xir 5th February 2013, 13:48 Quote
Prices should be coming down, we're not down from the floodrelated pricehike yet.
RichCreedy 5th February 2013, 13:48 Quote
if you are Intel, your storage is now mostly SSD based
SlowMotionSuicide 5th February 2013, 14:00 Quote
Quote:
Originally Posted by Mankz
The prices haven't fallen since the floods back to their original levels, while the prices and speeds of SSD's continue to fall to keep making them more attractive options.. that and more cloud based storage.

This.

My retailer of choice sold 1TB Spinpoint for 44,90 EUR just before the floods.
Now the cheapest comparable drive goes for 81,00 EUR. Plenty of headroom to drop prices.

I do know I'd get better EUR/GB ratio from bigger drives, though. Still gonna sit tight and wait for prices to drop.
Gareth Halfacree 5th February 2013, 14:38 Quote
Quote:
Originally Posted by theshadow2001
True. But, where as before 5 users would have 5 hard drives, with cloud storage 5 users could share a single cloud hard drive. Its a more efficient use of a single disk.
Not exactly: most cloud services are used in addition to local storage, rather than as a replacement for it. Take Dropbox, for example: you can't actually install the Dropbox client without it synchronising your files to your local PC.

Where that isn't the case is in mobile, where cloud storage is used instead of local storage - as with Google Music - but for desktops and laptops, the increasing popularity of cloud storage means more hard drive sales (the hard drive each user has anyway, plus a smaller number of larger hard drives for the datacentre) rather than fewer.
CampGareth 5th February 2013, 15:39 Quote
I'm still buying hard drives, until SSDs match them for price per GB they're gonna be useful. I mean I can't afford to pay for 3 512GB SSDs in my laptop, but I'll take 1 240GB SSD and a 1TB HDD. Oh and in my NAS 3TB drives are great, but then I store a lot of video, I'm a hoarder of sorts.
Ayrto 5th February 2013, 16:09 Quote
There's been consolidation since the floods, so will prices come down again? Not saying you've got cartel like price practices going on but 2TB drives could be had for as low as £46 pre- floods. Now they start at around £70
SchizoFrog 5th February 2013, 16:34 Quote
I think everyone is waiting for prices to come down. I have 5 physical drives in my current PC, 1x 1.5TB, 2x2TB, 1x 500GB and 1x 250GB. I am waiting for to come down so I can fill a NAS caddie with 3TB or 4TB drives for a decent price and then have an SSD in my main system. At the current prices it is just too expensive to justify the move as I still have a little space on my current setup.
Evolutionsic 5th February 2013, 16:35 Quote
Quote:
Originally Posted by Ayrto
There's been consolidation since the floods, so will prices come down again? Not saying you've got cartel like price practices going on but 2TB drives could be had for as low as £46 pre- floods. Now they start at around £70

2TB for £46? jesus how did i miss that!
theshadow2001 5th February 2013, 16:38 Quote
Quote:
Originally Posted by Gareth Halfacree
Not exactly: most cloud services are used in addition to local storage, rather than as a replacement for it. Take Dropbox, for example: you can't actually install the Dropbox client without it synchronising your files to your local PC.

Where that isn't the case is in mobile, where cloud storage is used instead of local storage - as with Google Music - but for desktops and laptops, the increasing popularity of cloud storage means more hard drive sales (the hard drive each user has anyway, plus a smaller number of larger hard drives for the datacentre) rather than fewer.

oops I forgot about the local storage aspect of the cloud
Ayrto 5th February 2013, 17:36 Quote
Quote:
Originally Posted by Evolutionsic
2TB for £46? jesus how did i miss that!

They have come down quite a lot from their post-flood peak . But they're still at least 40 percent higher than before the flood.
Harlequin 5th February 2013, 17:51 Quote
ssd prices are actually climbing- pre xmas a crucial M4 could be had for £100 for a 256 ; now they are £150 and climbing.


but , hdd`s do need to drop in price - there`s no supply issues yet the prices are sky high , ergo punters wont buy them.
south side sammy 5th February 2013, 18:38 Quote
Last drive I purchased was right after the flood. Knew I would need one and knew they would be hard to get because of the flood and because of the panic. Price was 60 bucks, 20 more than I usually paid. They're still priced high. Why would I even think about buying one until the price comes back to something more sensible? I don't feel sorry for the market. I held off purchasing 6 drives in the past year. I took some I had and wiped them where I normally would have bought new.
IvanIvanovich 5th February 2013, 18:43 Quote
The thing I find most perplexing is that they still bother to make drives smaller than 1TB. WHY? Especially when they are often the same price, or very close. For example I was looking recently and 500GB was same price as 2TB. Seems like they could lower cost by not offering so many models that I would guess there is little need for.
ZERO <ibis> 5th February 2013, 20:10 Quote
Quote:
Originally Posted by sniperdude
I want drives but I will wait till the prices drop

Exactly, I have been holding off on more drives to expand an array for a few years now. To save the money I just increased the amount of high compression 7z files to make more space. (for the type of files I store it can reduce space by >75%).

As I see the price of SSDs drop I am getting to a point where I might rather spend more to just get that b/c I know it is an actual value for my money. The HDD market will continue to feel like a ripoff to me until I see new low prices not simply a return to how they were 5 years ago.
ArthurB 5th February 2013, 21:58 Quote
Quote:
Originally Posted by lysol
Seems like they could lower cost by not offering so many models that I would guess there is little need for.
+1. People would also benefit from the higher areal density on the larger drives too.
yougotkicked 6th February 2013, 01:32 Quote
I hardly think the dip in price will actually hurt the R & D labs much; if they fall behind the moor's-law curve SSD's will just dominate the market faster. I for one think HDD's still have a long life ahead of them in the PC market. The technology has room to grow, and the data-center market is growing fast. The way I see it, the proliferation of mobile devices like tablets and smart phones, actually increases the amount of data the average person needs to store, thanks to embedded cameras, a constant stream of new apps, and the tendency to duplicate music collections across devices.

As for optical drives, I expect they will die out at about the same rate as the HTPC and smart TV's catch on, slowed of course by the media industry's steadfast opposition to progress, think floppy drives but with anyone at all saying they should stay around.
Xir 6th February 2013, 07:10 Quote
I wouldn't be so pessimistic for optical drives.
How long does it take to press the ~50GB of information on a blueraydisk through your average cheap DSL-line?
Same goes for streaming media, sure you can watch movies online, but if you go for high-quality full HD your looking at ~5-10GB even when compressed.

The same goes for harddisks.
Remember the "Retina" display? resolutions are going to silly levels, levels that will require incredible amounts of data. Imagine a pixeldensity like that (or like the Nexus 10's) on a large screen.
We don't even film on these resolutions yet, but boy are movies going to become storage eaters.
Gareth Halfacree 6th February 2013, 09:07 Quote
Quote:
Originally Posted by yougotkicked
I hardly think the dip in price will actually hurt the R & D labs much; if they fall behind the moor's-law curve SSD's will just dominate the market faster.
Moore's Law was an observation that the number of transistors in an integrated circuit doubles roughly every two years; it has very little impact on magnetic media, which is not an integrated circuit.
Maki role 6th February 2013, 12:25 Quote
I wonder if they're also simply pushing themselves out of the consumer market by releasing such massive drives? Most people don't need terabytes upon terabytes of storage, and they aren't reaching the limits of what they have. Sure there are many people who do need that storage (those with massive video and steam libraries for instance), but I wonder if people are just reaching what they need, much like I swear is happening with CPUs and in the tablet takeover. Could prove interesting a few years down the line though when standards like 4k hit the mainstream markets.
Ending Credits 6th February 2013, 16:41 Quote
Quote:
Originally Posted by Gareth Halfacree
Moore's Law was an observation that the number of transistors in an integrated circuit doubles roughly every two years; it has very little impact on magnetic media, which is not an integrated circuit.

Yes, but SSDs are governed by it, hence why magnetic media would need to match it to keep up.
Gareth Halfacree 6th February 2013, 17:34 Quote
Quote:
Originally Posted by Ending Credits
Yes, but SSDs are governed by it, hence why magnetic media would need to match it to keep up.
Not with the current price-per-gigabyte disparity: SSDs are way, way too expensive for bulk storage. Hard drive technology can sit on its hiney for the next decade and SSDs will still struggle to catch up. Let's assume SSDs match Moore's Law exactly: every two years, the transistor count doubles. Let's further simply things by assuming that means prices halve every two years. They won't, but it makes things nice and straightforward.

Currently, a 2TB mechanical hard drive will cost you between £50 and £70. A 2TB SSD... Well, there aren't any on Google Shopping, so we'll go for two 1TB models at a cost of around £2,000 each. So, every two years the cost of the SSD halves while the hard drive - in our simplified projection - stays the same because the industry is greedy, lazy or both. In two years' time, the SSDs cost £2,000 in total while the hard drive still costs £50. In four, £1,000. In six, £500. In eight, £250. In ten, £125. In twelve, £62.50. In fourteen years, £31.25.

So, it takes nearly a decade and a half for the cost of the 2TBs of SSD storage to beat that of the mechanical drive - and that's assuming the spinning-platter manufacturers don't increase areal density again or find some other way of increasing capacity and decreasing costs in that time period, which is pretty unlikely.

So, do hard drive manufacturers need to worry about Moore's Law? No, not really. At least, not this decade.

(Also, nothing is 'governed' by Moore's Law. It's a historical observation, nothing more. Okay, it's a historical observation that has proven surprisingly enduring and accurate, but it's an observation nevertheless.)
Cthippo 6th February 2013, 22:09 Quote
We're shooting on average 10-20 GB of video every day we go paddle, which is at least once a week, so, yeah, going to be saving on spinning platters for a while yet. I'm sure there is some kind of batch transcoding I could to do save space and get them all into one file format, but I haven't looked into that yet.

I think when we get back from California it's going to be time to start working on a new NAS box.
IvanIvanovich 7th February 2013, 00:10 Quote
Actually, those that still actually FILM, shooting at equivilent about 96MP resolution since oh the last 100 years.
fluxtatic 7th February 2013, 06:42 Quote
I picked up a 2TB Spinpoint for $80 not terribly long before the floods. I just checked Newegg, and damn near every 2TB is $110 and up. Mind, this was when the Spinpoints were real Samsung Spinpoints. I haven't checked (because prices are too high), but it used to be no-second-thought. New hard drive? Get a Spinpoint...I'm not certain that would still be the case anymore.

I've got 4TB in my NAS and a 500GB secondary in my desktop, so I'm good for a while. Eventually I'd like to fill the NAS with 3TB or 4TB drives, but not anytime soon, apparently. It was announce way, way back that capacity is back up to pre-flood levels, so now they're just profit mongering. Screw 'em.

As to optical drives - the internal in my desktop now is my last. I'll pick up a cheap USB external at some point, but I'm long past the point of needing an internal drive. I burn backups rarely, and the last time I've used it in the past year was for DVDs that I hadn't gotten around to ripping or torrenting.
mdshann 7th February 2013, 20:14 Quote
Quote:
Originally Posted by Mankz
The prices haven't fallen since the floods back to their original levels, while the prices and speeds of SSD's continue to fall to keep making them more attractive options.. that and more cloud based storage.

They are almost there though, I saw last week an ad for $59.99 1TB hitachis.
yougotkicked 10th February 2013, 04:14 Quote
Quote:
Originally Posted by Gareth Halfacree
Not with the current price-per-gigabyte disparity: SSDs are way, way too expensive for bulk storage. Hard drive technology can sit on its hiney for the next decade and SSDs will still struggle to catch up.

...

(Also, nothing is 'governed' by Moore's Law. It's a historical observation, nothing more. Okay, it's a historical observation that has proven surprisingly enduring and accurate, but it's an observation nevertheless.)

I concede that I misapplied Moor's law, but after a little poking about i came across Kryder's Law, described by one of the top-dogs at Seagate a few years back. Statistically speaking, magnetic storage technology has advances with the same exponential trend as transistor density, but faster.

Based on a statistical analysis I did for a paper in highschool, I'd venture to say the the industry is governed by Moore's Law, I don't have a digital copy of the paper anymore, so I don't recall the specifics, but I found that the correlation between flagship processors' transistor count and numbers projected by moor's law relative the the flagship processor when he made the first assertion, was over 99%. Any statistician will tell you that you don't see correlations that strong by chance. The entire industry had a formula with which they could estimate the performance level of their competitors next product, so the whole industry set their pace by it.

Just some food for thought.
theshadow2001 10th February 2013, 14:28 Quote
Wasn't moore's law more of a self fulfilling prophecy which was based on an observation.
yougotkicked 10th February 2013, 22:27 Quote
In a sense yes. Gordon Moore made the observation in 1965 based on development trends he observed at Intel. However, the prediction has proved accurate for almost 6 decades, across multiple breakthroughs in transistor technology, and for companies that have no affiliation with Intel's R&D. An argument can be made that the industry was influenced by the prediction, but I wouldn't go as far as to say that it's a self fulfilling prophecy.
theshadow2001 10th February 2013, 23:30 Quote
Quote:
Originally Posted by yougotkicked
In a sense yes. Gordon Moore made the observation in 1965 based on development trends he observed at Intel. However, the prediction has proved accurate for almost 6 decades, across multiple breakthroughs in transistor technology, and for companies that have no affiliation with Intel's R&D. An argument can be made that the industry was influenced by the prediction, but I wouldn't go as far as to say that it's a self fulfilling prophecy.

Well if the industry was influenced by his prediction and they matched what he predicted then surely by any accounts its a self fulfilling prophecy. You can replace prophecy with prediction to make it seem less mystical or wizardy.
yougotkicked 11th February 2013, 00:18 Quote
well, my hold-up with calling it a self-fulfilling prophecy is that it was by no means a trivial thing for the industry to match his predictions. It isn't a question of companies choosing to meet those predictions, moore's law essentially states that every 2 years, companies will figure out a way to make chips 2 times better than they were 2 years ago.

What makes moore's law special is not simply that his equation was accurate, it's special because of what that equation entails. No other sort of technology in the history of man kind has experienced such rapid growth for such a long time, and he made the prediction not too long into the life of the transistor. If cars had improved at a similar rate over the last 50 years, you could get a million miles to the gallon in a car you bought with loose change you found in the couch. If it weren't true, the prediction would have made him look like a babbling idiot.
Gareth Halfacree 11th February 2013, 08:10 Quote
Quote:
Originally Posted by yougotkicked
Based on a statistical analysis I did for a paper in highschool, I'd venture to say the the industry is governed by Moore's Law, I don't have a digital copy of the paper anymore, so I don't recall the specifics, but I found that the correlation between flagship processors' transistor count and numbers projected by moor's law relative the the flagship processor when he made the first assertion, was over 99%. Any statistician will tell you that you don't see correlations that strong by chance. The entire industry had a formula with which they could estimate the performance level of their competitors next product, so the whole industry set their pace by it.
Congratulations, you've fallen for Intel's re-spinning of the story. Gordon Moore never made any such prediction: he made an observation about what had gone in the past. In other words: transistor counts had been roughly doubling every two years for some time, and Moore was merely the first to put that fact down on paper.

Don't believe me? Here's a handy PDF of the article in which 'Moore's Law' was first codified. Far from being a prediction, or even an equation, here's the 'law' in full:
Quote:
Originally Posted by Gordon Moore
The complexity for minimum component costs has increased at a rate of roughly a factor of two per year.
That's it. That's the entire law, in full: an observation that, historically, the complexity of a semiconductor has increased at a rate that sees it roughly doubling every two years. No predictions, no "strong correlation," no equations: a simply-worded and very rough historical observation.

Moore does venture into the realm of prediction in that paragraph, but still in a very woolly manner - and phrased as "basically, nothing will change for the next decade, unless it does."
Quote:
Originally Posted by Gordon Moore
Certainly over the short term this rate can be expected to continue, if not to increase. Over the longer term, the rate of increase is a bit more uncertain, although there is no reason to believe it will not remain nearly constant for at least 10 years. That means by 1975, the number of components per integrated circuit for minimum cost will be 65,000.

If you can find a 'formula' in that which shows '99 per cent' correlation to actual transistor counts, congratulations: a bright future awaits you in a council planning department.
yougotkicked 13th February 2013, 07:04 Quote
Quote:
Originally Posted by Gareth Halfacree
Congratulations, you've fallen for Intel's re-spinning of the story. Gordon Moore never made any such prediction: he made an observation about what had gone in the past. In other words: transistor counts had been roughly doubling every two years for some time, and Moore was merely the first to put that fact down on paper.

Don't believe me? Here's a handy PDF of the article in which 'Moore's Law' was first codified. Far from being a prediction, or even an equation, here's the 'law' in full:
That's it. That's the entire law, in full: an observation that, historically, the complexity of a semiconductor has increased at a rate that sees it roughly doubling every two years. No predictions, no "strong correlation," no equations: a simply-worded and very rough historical observation.

Moore does venture into the realm of prediction in that paragraph, but still in a very woolly manner - and phrased as "basically, nothing will change for the next decade, unless it does."

If you can find a 'formula' in that which shows '99 per cent' correlation to actual transistor counts, congratulations: a bright future awaits you in a council planning department.

Wow, people are being way more hostile to me lately than I think is reasonable. Maybe my memory is failing me and I actually did post a bunch of inflammatory remarks earlier.

I apologize for associating the man who the principle is named for with the slightly more specific connotations said principle.

I actually read the entire article where Moore's Law was first described years back when I was writing a paper on it, I just didn't take such a cynical view of the loose relationship between his statements and the popularized version of Moore's Law. Perhaps because it was not named that by a team of scheming Intel publicists; the term was coined by Carver Mead, Professor Emeritus at Caltech. Or, possibly because it has been called Moore's Law for a long time, everyone knows it by that name, and Dr. Moor is a great man and a pioneer of the industry.

I DID find a formula for 99 percent correlation, for a mandatory research paper in a high school statistics course. I codified the general form of the law (doubling every 2 years) and set its y-intercept to the transistor count of Intel's flagship processor in 1965. I then gathered data points from other flagship processors over the years to create a fit curve. I then used the Chi-squared distribution test (after shifting both equations to a logarithmic scale) to find the correlation coefficient between Moore's 1965 statement and the actual growth of technology since then. The result was an unnaturally high coefficient of 0.99....

I think I was 15 at the time. Currently I'm earning a degree in computer science and working as a web developer, IT rep, and programming consultant for research faculty; where I'm currently helping researchers in the bio-medical engineering department leverage GPGPU computing in one of their more time-consuming simulations.

I was planning on doing something with that in the future, but if you think my afternoon's work all those years ago means I should reconsider my plans, I have some serious thinking to do.
Gareth Halfacree 13th February 2013, 07:58 Quote
Quote:
Originally Posted by yougotkicked
Wow, people are being way more hostile to me lately than I think is reasonable.
I apologise; my posts were not meant to come across as hostile, but re-reading them - certainly the most recent, anyway - I can see how they could be interpreted as such. Mea maxima culpa.
yougotkicked 14th February 2013, 01:35 Quote
I'm sorry too, I over-reacted in my last post. I've been a bit over-loaded lately, and for the past week it seems like I've had to defend every little thing I say on the internet.

I was irritated already; you made some unfortunate choices while phrasing your post.

Forgiven and forgotten.
[USRF]Obiwan 11th April 2013, 11:42 Quote
Since the 'flood-prices' I just delete (read; HD space management) content from my three 1.5Gb drives. Instead of being lazy and leave it sit for 1000 years and add more and bigger drives for my 'collecting rage'.
In turn this altered my brain to spend more wisely...
NethLyn 14th April 2013, 15:46 Quote
WD have kept the price of the WD Black high at £76 for the 1TB, with the introduction of this Red variety, if their variable spin speed in the Green and Red drives makes them slower than usual for gaming then even if they're cheaper, I'll carry on holding out for the Black to come down.

It's taken between the first floods and now to finish the burning and mass deletion of all my backups of the past six years, now I just try not to download so much junk to have that problem again.
Log in

You are not logged in, please login with your forum account below. If you don't already have an account please register to start contributing.



Discuss in the forums