Me too. I have too much invested in SanDisk. I guess we all hope that WD doesn't mess up a good thing. Smart move by WD. I'm not sure I want one's success to depend on the other.
I don't like the over-dramatization I've seen in some of the AnandTech articles lately. This one acts as though hard drives are (or are becoming) niche storage. Hard drives are far from niche. Though they may be mostly removed from consumer computers within the next few years, hard drives will continue to have a very healthy market in the enterprise storage space (which is massive). Consumers will still be using them in the form of backup/external storage as well.
Enterprise would be niche and even worse their ASP in enterprise will decline too.. Anyway, you don't know the size of that market. WD actually provides non-PC revenue and you can do the math and approximate enterprise. In Q2 enterprise was 15% of their units( 7.2 million out of 48.5 - they provide those numbers) and enterprise revenue was likely some 1.15 billions so some 36% of revenue. Even factoring in some variations and accepting a 35-40% of revenue range ,that not that much. Losing scale , by losing all other segments almost entirely would have a huge impact on costs and they would be in far more trouble than you imagine. Being the new tape isn't a great business.
Honestly... they kind of are. IMO 2-3 years SSD's prices should be down close enough to HDD as far as price/gb as to be irrelevant. Right now the biggest complaint i encounter in my various forums is that people want more space for their games. Basically a lot of people have slow internet, or caps on their internet, and like to keep 40 or 50 games installed at a time so they dont have to redownload it, etc, if they decide to play it.
Personally i think they're being ridiculous. I have this argument regularly with my best friend who still hasn't bought an SSD because he maintains that he needs to keep all these games installed, yet when i look at his steam profile, many of them he literally hasn't played 1 min in 1-2 years or longer.
However, to get back to my point, once you can get a 512gb SSD for $100 or so or less, i personally (and this is just my anecdotal opinion) think that the average gamer will switch over. Right now gamers occupy this weird middle ground in storage needs. The average user, i.e. web surfer, email, youtuber, etc, doesn't need multiple TB of storage, and if they do they typically use portable mediums such as external HDDs. People who really do need a ton of space (video editors, photo editors, other professionals, etc) will be the main source of people who need high capacity HDD's that aren't *as* concerned about performance of the drive.
So, in my humble opinion, once you convert gamers over to SSD primarily, you will have basically reduced the HDD market to professionals, and things like NAS/SAN/etc.
I like having both. Like my main notebook I've got a 500GB Micron SSD + 750GB 7200RPM mechanical drive. I bought a 1TB SSD for my newer though lower end system, but still, I like having the OS and whatnot on an SSD for faster boot times, and I keep my media including games on my mechanical drive.
Outside of the enthusiast market, I think you're being overly optimistic. $400 race to the bottom laptops sell in much larger volumes than anything else. As long as they can save $0.01 by using an HDD over an SSD they will.
You're also IMO underestimating the value of bigger bogo-stats to non-technical buyers. Latency/random IO rate/IOPS are completely meaningless to most people. 1000 GB means I can store more than 8 times as much stuff at 120 GB is something that they can intuitively understand and "know" that the big HDD is better than the small SSD. Besides 120GB would be worse than the 320GB in their last laptop, and you can't buy a new laptop that's worse than what you're replacing. (Even though 99% of them won't actually use any of the extra space).
Unless cheap ultrabooks kill off cheap conventional thickness worstbuy specials, HDDs will continue to dominate consumer sales until the sticker price per GB is lower for SSDs; which probably won't be for a number of years. (The articles claiming the SSD cost/GB will cross over the HDD one for bulk storage in the next year or two are looking at TCO for datacenter use, where savings in power use and server density are much more important than the upfront price that dominates cheap laptop specs.)
The question is, though, whether any company really wants to be in or bet on the low end market segment product lines. WD would have to sell a shit load of drives in $400 laptops in order to make any kind of significant profit. So unless the profit margins on big HDDs are larger than they are on SSDs, HDDs are going to go out of style (especially with more and more storage needs being covered by various cloud services). Perhaps not in 2-3 years, but WD has to think 5+ years, and that’s a realistic timeframe for their HDD business to become a niche venture.
Until SSDs become cheaper per GB for bulk (consumer) and semi/offline (enterprise) data storage a market for high capacity 3.5" HDDs will still be around. In the medium term, the enterprise portion of the market should be big enough to sustain the R&D costs which then trickle down to mass market consumer devices like your worst buy specials (which have never been priced at a level to recover significant R&D money). The hollowing out of the middle of the market (HDDs in mid/high end laptops/desktops and general purpose servers); will slow the rate of R&D down; but give or take what would've been the final HDD generation before SSDs overtake in dollars/gb for cheap storage (which may end up being axed out of fear that it might not break even before being obsoleted) I'd expect R&D to continue until the transition point with the HDD companies continuing to churn out the last model they designed until OEM design finally vanishes.
If the CEO of any one of the HDD company think like that I would not own their Stock. Enterprise Storage is NOT massive. As @jjj have stated, and it wont grown forever. The increase in Enterprise unit will properly never make up for the lost of consumer electronics. It is also worth noting 3D NAND has a very bright future ahead, so far they do not see a ceiling in 3D NAND stacks, with roadmap of up to 512 layers, we may see 1024 layers ( we dont know, this is likely 8 years down the road ) and that does not include smaller nodes or Quad Level Cell to increase capacity. If you were worrying about the P/E cycles, as the larger the NAND storage, the less likely you will reach that limit in day to day usage.
So we have ~500x capacity increase within the next 10 years. That is roughly doubling the capacity every year.
Today there are still many web servers using HDD. but in 2020 we will likely see even Web Servers are all SSD.
Why does NAND have to move so fast? You may ask, they could certainly milk the industry together, but remember there is Xpoint. Those expensive, and high profitable Enterprise SSD storage wont last forever. Xpoint provides everything Enterprise want and are willing to pay for it.
Not. Companies look forward into the future as far as they can and spinning drives will become a niche as tape drives did. SSDs will have more density than HDDs in the near future and price per GB will approach HDDs a few years after that. That excludes the lower power, reliability, and much higher performance.
SSDs don't necessarily have lower power, and I'm REALLY not sure about any better reliability. Flash storage seems pretty flaky compared to mechanical disks so far...
It seems the industry has by and large embraced SSDs around when Intel pushed out the S3700 (and S3600/S3500), judging by the somewhat recent flood of them on ebay. Partly it comes from OEMs having them in their server config setup, and partly from various testbeds left, right and center proving quite reliable. In fact, the issues in general seem to be from bad OS code (850EVO/Pro) or a bad controller . Of course, exceptions like the SF-2281 failures and 840/840EVO mar things a bit, but across the industry, especially the server space, things have been pretty quiet... Or I'm reading the wrong forums...
At the other extreme, SSDs are absolutely miles ahead of HDDs in any sort of mobile use simply by being shockproof.. so far I've killed 2 2.5" disks, and have another 2 that are rather suspect (had some data loss, not sure if software bug or shock)
And finally, for lower power, outside of something like an S3500 that's distinctly at the extreme high-end with 4.27W power consumption, laptop HDDs are almost universally higher power the moment they spin up, all the while having massively better performance.
"Toshiba and SanDisk are lagging behind in the transition to 3D NAND"
That's nonsense. The transitions are only about the financials. Samsung is said to have very high costs and very little 3D production while shipping only in consumer since nothing is qualified in enterprise. Tosh/Sandisk are still ramping 15nm (Sandisk said today 60% was 15nm in Q3) and ,as you say, they lead there so why spend a lot of money to transition early? Besides who else has 3D in significant volumes , who else has 48 layers? Micron is only now ramping 2D TLC lol , Sandisk is behind because they wouldn't waste money to push 3D early? In the controller space the software is more important than the actual chip.
The far trickier transition is to RRAM or other new memory and it will have to be soon enough but there they got their own research and their partners, Toshiba and the new alliance with HP.
Even a blind squirrel finds a nut once in a while ;-).
Honestly, i'm just ribbing you. It does feel nice to have a prediction come true though. My most satisfying was predicting the dismal failure that is Wildstar.
All meaning the newly released SM863 and PM863, those are too new to matter and you are deflecting. By the end of next year maybe 20-25% of wafer output will be 3D and chances are Toshiba/Sandisk will have the best or the second best solution. Micron/Intel are first pushing MLC but the others will go TLC in SSD , mobile and removable. They went floating gate and they'll be a bit slower because of it and ofc they got fewer layers. Hynix seems a bit fuzzy on timing and volumes so chances are they got cost problems for now. We'll see what Samsung does but they might need a shrink to beat Toshiba/Sandisk.
Out of the branded enterprise SSDs, the previous gen 845DC PRO also used V-NAND (24-layer MLC). If you look at the OEM SAS and PCIe models you will find quite a few products using V-NAND and only a couple of older products with planar NAND.
There is no XY shrink planned for 3D nand, all the manufacturers agreed during the Flash Memory Summit keynote this year. Shrinking XY would risk toppling the Z structure. Everyone is talking about higher and higher Z layer count though.
I don't think hard drives are going away any time soon either.
At home, on my main PC, I have one SSD for the OS and some games, mostly ones that benefit the most from quick access storage and that I'm playing frequently still, two HDDs for media (movies, music, etc) and one extra HDD where most of my Steam repository stays (infrequent played games, etc).
There's no way I could fit all that data on SSDs, the cost would be too much, and gains minimal (media reading... media recording is another story).
Perhaps most people don't need more than their 512GB/1TB SSD on their notebooks, or something. That's not my case. I certainly could put all the media into a NAS or fileserver, but the games would still have to stay within my machines.
As HDD sizes grow tho, it becomes a bit scary the amount of data you can lose if all your fault-tolerances end up failing. =/
Then there are just cheapskates like myself who will buy 2TB in HDD form because it's cheaper than the SSD and kind of works.
I've often wondered how long it will take SSDs to replace 90% of HDDs but I think it's 5 years away at least. When I replace my 2013 PC - and that's 2018 or later with the current slow rate of progress in CPUs - I may get an SSD by default because it comes with a new system, but no earlier than that.
But the point is you don't NEED an SSD to store a bunch of videos/music, etc. Where the benefits of an SSD are apparent are OS drives, and game drives. Now, i will say that there is some argument to the issue of gaming, particularly with monstrosities like BF4 at 56gb and TW3 at 50+ GB, etc. However, i think we're not too far, 2-3 years at most, from being able to get 512gb to 1TB SSD's at a reasonable cost ($100-150) and at that point the argument for gaming starts to go out the window IMO.
I think too many people are looking at whatever their current total storage is and equating them to having to replace all of that with SSD. I think if people prioritized (like you did) they could reap the benefits of an SSD and be perfectly fine on storage. SSD for a gaming drive and particularly on your OS drive is a night and day difference. It was far and away the best quality of life decision i ever made on my PC.
I kinda feel that peeps installing all of their games on an SSD are only doing it for emotional reasons, the difference in-game is minimal.
The reason I like SSDs, especially in laptops, is they don't suffer from being shaken/shocked like a HDD does. Same reason why I don't use external drives anymore, when high capacity USB / uSD cards are available.
Until networking (home network and internet providers) gets to a stage where your HDD is the limitation, I can't see a big movement over to SSDs until it becomes cost effective to.
Crappy networks are more a barrier to progress than SSDs/HDDs, reliable 4G/5G networks rolled out for 100% coverage with enough resources put in to give unlimited data for all uses (mobiles/tablet/desktop/home server). Surely this is cheaper than taking a fibre optic cable to each and every house?
4g is only as fast as it is because there are bandwidth caps. People aren't apeshit pounding the network like they do with their land lines. If you had as many users as landlines do, all using it as much as they would normally use their land lines, it would absolutely decimate those networks. And really, you don't need fiber to each house. Coax is more than capable of stupidly fast internet speeds. The issue is getting the cable/telecom companies to not have monopoly access to local municipalities. In all the places in the US where there is actual competition, speeds are great on cable and prices are reasonable. In the places where the useless city council have given monopoly access to the cable infrastructure to one company, its a crap shoot.
But seriously, i can get a 250/50 connection on coax right now. Its ungodly expensive, but it can be done. Fiber is NOT necessary.
Speeds here in Washington state have been increasing rather quickly in the past year or so. Most of the surrounding state especially Seattle suburbs only have access to Comcast yet comcast has been steadily increasing speeds. At the end of last year, 2014, comcast increased our home speed from 50mps to 105mbps for the same cost per month. After 5 or 6 months of having 105mps I called comcast and asked to be downgraded to 50mbps to save some money. Now again, just this month, they upped our speed from 50 to 75mbps, and I'm fairly certain if I had stayed at 105 that that package was increased to 150mbps. To be honest I don't believe most people even need speeds over 50mbps, I sure don't. It doesnt really affect me much whether my downloading speeds are at 6MB/s or 10MB/s, I still enjoy the same quality of streaming and internet usage, and my household usage on average is about 400GB/month. Also all of the these speed tiers are fairly affordable especially when they give you promotional rates for 6 months to a year and even at the end of the promotion if you call they usually, without much fuss at all, extend your promotional period. Also it only costs an extra $10! dollars per month to get bumped up to the next speed tier, which used to mean going from 50 to 105 in my case a few months ago...I'm not sure what the diff speed tiers are now, now that they have increased their speeds again. My purpose for this post was in reference to all the comments about Google fiber, and gigabit speeds, and such that I always see... I don't think most private residents need anywhere close to gigabit speeds. I will agreee that businesses are a completely diff story and yes there are special cases out there whom I would consider the enthusiasts/power users (including online gamers who want low latency but again they are in the minority). Even with just 75mbps here at home I show about a 9 to 10ms ping consistently, even across large distances the ping is fairly low. Also one last thing comcast is great at always delivering speeds that are always better than what you would expect. With the 50mbps package I always averaged somewhere in the 60s(mbp/s). Now having 75mbps I avg around 85-90. In No Way am I pro Comcast but I think internet speeds, especially in my area and I would guess in most urban areas, are to the point where most private residents should be satisfied. Also even without any competition comcast still continues to upgrade their speeds around here. I can easily subscribe to speeds upto 250mbps for about $150/month and in my opinion if you need speeds like that then that price point shouldn't be too far out of reach. You can easily get speeds of 150mbps for less than $100 around here. I am not naive and do understand that not all areas,especially around the world, have access to decent speeds but I have seen great progress being made in the past couple years, in the US. Would I love gigabit fiber, of course...Do I 'Need' gigabit fiber, if I'm being practical, of course not and I would argue that that would apply to most Americans at least for the immediate future.
well in most places in the world speeds per $ monthly aren't progressing as fast as storage capabilities and file size. For example where I live I get 20 mbit but for like 15 years I've not seen any free upgrade or anything. And I'm in one of the fastest countries in western Europe.
If you take countries like Romania this problem doesn't exist of course. That's just because a country with not even 10 million people has hundreds of ISPs. Other countries are bogged down by the inheritance of state monopolies.
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
35 Comments
Back to Article
MadHatter0 - Wednesday, October 21, 2015 - link
Here's hoping the quality and warranty on Sandisk under it's new owner holds up!bigboxes - Friday, October 23, 2015 - link
Me too. I have too much invested in SanDisk. I guess we all hope that WD doesn't mess up a good thing. Smart move by WD. I'm not sure I want one's success to depend on the other.vmll - Wednesday, October 21, 2015 - link
I don't like the over-dramatization I've seen in some of the AnandTech articles lately. This one acts as though hard drives are (or are becoming) niche storage. Hard drives are far from niche. Though they may be mostly removed from consumer computers within the next few years, hard drives will continue to have a very healthy market in the enterprise storage space (which is massive). Consumers will still be using them in the form of backup/external storage as well.jjj - Wednesday, October 21, 2015 - link
Enterprise would be niche and even worse their ASP in enterprise will decline too.. Anyway, you don't know the size of that market.WD actually provides non-PC revenue and you can do the math and approximate enterprise. In Q2 enterprise was 15% of their units( 7.2 million out of 48.5 - they provide those numbers) and enterprise revenue was likely some 1.15 billions so some 36% of revenue. Even factoring in some variations and accepting a 35-40% of revenue range ,that not that much. Losing scale , by losing all other segments almost entirely would have a huge impact on costs and they would be in far more trouble than you imagine. Being the new tape isn't a great business.
Kutark - Wednesday, October 21, 2015 - link
Honestly... they kind of are. IMO 2-3 years SSD's prices should be down close enough to HDD as far as price/gb as to be irrelevant. Right now the biggest complaint i encounter in my various forums is that people want more space for their games. Basically a lot of people have slow internet, or caps on their internet, and like to keep 40 or 50 games installed at a time so they dont have to redownload it, etc, if they decide to play it.Personally i think they're being ridiculous. I have this argument regularly with my best friend who still hasn't bought an SSD because he maintains that he needs to keep all these games installed, yet when i look at his steam profile, many of them he literally hasn't played 1 min in 1-2 years or longer.
However, to get back to my point, once you can get a 512gb SSD for $100 or so or less, i personally (and this is just my anecdotal opinion) think that the average gamer will switch over. Right now gamers occupy this weird middle ground in storage needs. The average user, i.e. web surfer, email, youtuber, etc, doesn't need multiple TB of storage, and if they do they typically use portable mediums such as external HDDs. People who really do need a ton of space (video editors, photo editors, other professionals, etc) will be the main source of people who need high capacity HDD's that aren't *as* concerned about performance of the drive.
So, in my humble opinion, once you convert gamers over to SSD primarily, you will have basically reduced the HDD market to professionals, and things like NAS/SAN/etc.
dsumanik - Wednesday, October 21, 2015 - link
i disagree flipping point will be, 1tb for 100 or less.Wolfpup - Thursday, October 22, 2015 - link
I like having both. Like my main notebook I've got a 500GB Micron SSD + 750GB 7200RPM mechanical drive. I bought a 1TB SSD for my newer though lower end system, but still, I like having the OS and whatnot on an SSD for faster boot times, and I keep my media including games on my mechanical drive.DanNeely - Thursday, October 22, 2015 - link
Outside of the enthusiast market, I think you're being overly optimistic. $400 race to the bottom laptops sell in much larger volumes than anything else. As long as they can save $0.01 by using an HDD over an SSD they will.You're also IMO underestimating the value of bigger bogo-stats to non-technical buyers. Latency/random IO rate/IOPS are completely meaningless to most people. 1000 GB means I can store more than 8 times as much stuff at 120 GB is something that they can intuitively understand and "know" that the big HDD is better than the small SSD. Besides 120GB would be worse than the 320GB in their last laptop, and you can't buy a new laptop that's worse than what you're replacing. (Even though 99% of them won't actually use any of the extra space).
Unless cheap ultrabooks kill off cheap conventional thickness worstbuy specials, HDDs will continue to dominate consumer sales until the sticker price per GB is lower for SSDs; which probably won't be for a number of years. (The articles claiming the SSD cost/GB will cross over the HDD one for bulk storage in the next year or two are looking at TCO for datacenter use, where savings in power use and server density are much more important than the upfront price that dominates cheap laptop specs.)
Kutark - Friday, October 23, 2015 - link
You make excellent pointsxype - Monday, October 26, 2015 - link
The question is, though, whether any company really wants to be in or bet on the low end market segment product lines. WD would have to sell a shit load of drives in $400 laptops in order to make any kind of significant profit. So unless the profit margins on big HDDs are larger than they are on SSDs, HDDs are going to go out of style (especially with more and more storage needs being covered by various cloud services). Perhaps not in 2-3 years, but WD has to think 5+ years, and that’s a realistic timeframe for their HDD business to become a niche venture.DanNeely - Monday, October 26, 2015 - link
Until SSDs become cheaper per GB for bulk (consumer) and semi/offline (enterprise) data storage a market for high capacity 3.5" HDDs will still be around. In the medium term, the enterprise portion of the market should be big enough to sustain the R&D costs which then trickle down to mass market consumer devices like your worst buy specials (which have never been priced at a level to recover significant R&D money). The hollowing out of the middle of the market (HDDs in mid/high end laptops/desktops and general purpose servers); will slow the rate of R&D down; but give or take what would've been the final HDD generation before SSDs overtake in dollars/gb for cheap storage (which may end up being axed out of fear that it might not break even before being obsoleted) I'd expect R&D to continue until the transition point with the HDD companies continuing to churn out the last model they designed until OEM design finally vanishes.iwod - Wednesday, October 21, 2015 - link
If the CEO of any one of the HDD company think like that I would not own their Stock.Enterprise Storage is NOT massive. As @jjj have stated, and it wont grown forever. The increase in Enterprise unit will properly never make up for the lost of consumer electronics.
It is also worth noting 3D NAND has a very bright future ahead, so far they do not see a ceiling in 3D NAND stacks, with roadmap of up to 512 layers, we may see 1024 layers ( we dont know, this is likely 8 years down the road ) and that does not include smaller nodes or Quad Level Cell to increase capacity. If you were worrying about the P/E cycles, as the larger the NAND storage, the less likely you will reach that limit in day to day usage.
So we have ~500x capacity increase within the next 10 years. That is roughly doubling the capacity every year.
Today there are still many web servers using HDD. but in 2020 we will likely see even Web Servers are all SSD.
Why does NAND have to move so fast? You may ask, they could certainly milk the industry together, but remember there is Xpoint. Those expensive, and high profitable Enterprise SSD storage wont last forever. Xpoint provides everything Enterprise want and are willing to pay for it.
zodiacfml - Thursday, October 22, 2015 - link
Not. Companies look forward into the future as far as they can and spinning drives will become a niche as tape drives did. SSDs will have more density than HDDs in the near future and price per GB will approach HDDs a few years after that. That excludes the lower power, reliability, and much higher performance.Wolfpup - Thursday, October 22, 2015 - link
SSDs don't necessarily have lower power, and I'm REALLY not sure about any better reliability. Flash storage seems pretty flaky compared to mechanical disks so far...Wolfpup - Thursday, October 22, 2015 - link
EDIT: That said I've not had issues with my Micron/Crucial drives nor my Intel drives like I've had with other brands of flash.ZeDestructor - Thursday, October 22, 2015 - link
It seems the industry has by and large embraced SSDs around when Intel pushed out the S3700 (and S3600/S3500), judging by the somewhat recent flood of them on ebay. Partly it comes from OEMs having them in their server config setup, and partly from various testbeds left, right and center proving quite reliable. In fact, the issues in general seem to be from bad OS code (850EVO/Pro) or a bad controller . Of course, exceptions like the SF-2281 failures and 840/840EVO mar things a bit, but across the industry, especially the server space, things have been pretty quiet... Or I'm reading the wrong forums...At the other extreme, SSDs are absolutely miles ahead of HDDs in any sort of mobile use simply by being shockproof.. so far I've killed 2 2.5" disks, and have another 2 that are rather suspect (had some data loss, not sure if software bug or shock)
And finally, for lower power, outside of something like an S3500 that's distinctly at the extreme high-end with 4.27W power consumption, laptop HDDs are almost universally higher power the moment they spin up, all the while having massively better performance.
Michael Bay - Saturday, October 24, 2015 - link
What exactly waqs wrong with 850evo/pro? I`m aware of 840 issues, but not 850.beginner99 - Thursday, October 22, 2015 - link
> This one acts as though hard drives are (or are becoming) niche storage.They will sooner or later. Western Digital could to something about it or stay stubborn and go the way of Kodak.
jjj - Wednesday, October 21, 2015 - link
"Toshiba and SanDisk are lagging behind in the transition to 3D NAND"That's nonsense. The transitions are only about the financials. Samsung is said to have very high costs and very little 3D production while shipping only in consumer since nothing is qualified in enterprise. Tosh/Sandisk are still ramping 15nm (Sandisk said today 60% was 15nm in Q3) and ,as you say, they lead there so why spend a lot of money to transition early? Besides who else has 3D in significant volumes , who else has 48 layers? Micron is only now ramping 2D TLC lol , Sandisk is behind because they wouldn't waste money to push 3D early?
In the controller space the software is more important than the actual chip.
The far trickier transition is to RRAM or other new memory and it will have to be soon enough but there they got their own research and their partners, Toshiba and the new alliance with HP.
jjj - Wednesday, October 21, 2015 - link
BTW more than 2 years ago in comment here i was suggesting that WD should buy Sandisk http://www.anandtech.com/show/6943/wd-and-sandisk-...Kutark - Wednesday, October 21, 2015 - link
Even a blind squirrel finds a nut once in a while ;-).Honestly, i'm just ribbing you. It does feel nice to have a prediction come true though. My most satisfying was predicting the dismal failure that is Wildstar.
dsumanik - Wednesday, October 21, 2015 - link
this is just gonna mean more $$ for already overpriced sandisk storage, anyone who thinks otherwise is dreamingKristian Vättö - Thursday, October 22, 2015 - link
All Samsung's enterprise drives use V-NAND, which are used by a number of enterprise storage array vendors.jjj - Thursday, October 22, 2015 - link
All meaning the newly released SM863 and PM863, those are too new to matter and you are deflecting. By the end of next year maybe 20-25% of wafer output will be 3D and chances are Toshiba/Sandisk will have the best or the second best solution. Micron/Intel are first pushing MLC but the others will go TLC in SSD , mobile and removable. They went floating gate and they'll be a bit slower because of it and ofc they got fewer layers. Hynix seems a bit fuzzy on timing and volumes so chances are they got cost problems for now. We'll see what Samsung does but they might need a shrink to beat Toshiba/Sandisk.Kristian Vättö - Thursday, October 22, 2015 - link
Out of the branded enterprise SSDs, the previous gen 845DC PRO also used V-NAND (24-layer MLC). If you look at the OEM SAS and PCIe models you will find quite a few products using V-NAND and only a couple of older products with planar NAND.frenchy_2001 - Thursday, October 22, 2015 - link
There is no XY shrink planned for 3D nand, all the manufacturers agreed during the Flash Memory Summit keynote this year. Shrinking XY would risk toppling the Z structure.Everyone is talking about higher and higher Z layer count though.
LordanSS - Wednesday, October 21, 2015 - link
I don't think hard drives are going away any time soon either.At home, on my main PC, I have one SSD for the OS and some games, mostly ones that benefit the most from quick access storage and that I'm playing frequently still, two HDDs for media (movies, music, etc) and one extra HDD where most of my Steam repository stays (infrequent played games, etc).
There's no way I could fit all that data on SSDs, the cost would be too much, and gains minimal (media reading... media recording is another story).
Perhaps most people don't need more than their 512GB/1TB SSD on their notebooks, or something. That's not my case. I certainly could put all the media into a NAS or fileserver, but the games would still have to stay within my machines.
As HDD sizes grow tho, it becomes a bit scary the amount of data you can lose if all your fault-tolerances end up failing. =/
stephenbrooks - Wednesday, October 21, 2015 - link
Then there are just cheapskates like myself who will buy 2TB in HDD form because it's cheaper than the SSD and kind of works.I've often wondered how long it will take SSDs to replace 90% of HDDs but I think it's 5 years away at least. When I replace my 2013 PC - and that's 2018 or later with the current slow rate of progress in CPUs - I may get an SSD by default because it comes with a new system, but no earlier than that.
Kutark - Wednesday, October 21, 2015 - link
But the point is you don't NEED an SSD to store a bunch of videos/music, etc. Where the benefits of an SSD are apparent are OS drives, and game drives. Now, i will say that there is some argument to the issue of gaming, particularly with monstrosities like BF4 at 56gb and TW3 at 50+ GB, etc. However, i think we're not too far, 2-3 years at most, from being able to get 512gb to 1TB SSD's at a reasonable cost ($100-150) and at that point the argument for gaming starts to go out the window IMO.I think too many people are looking at whatever their current total storage is and equating them to having to replace all of that with SSD. I think if people prioritized (like you did) they could reap the benefits of an SSD and be perfectly fine on storage. SSD for a gaming drive and particularly on your OS drive is a night and day difference. It was far and away the best quality of life decision i ever made on my PC.
Robalov - Thursday, October 22, 2015 - link
I kinda feel that peeps installing all of their games on an SSD are only doing it for emotional reasons, the difference in-game is minimal.The reason I like SSDs, especially in laptops, is they don't suffer from being shaken/shocked like a HDD does. Same reason why I don't use external drives anymore, when high capacity USB / uSD cards are available.
Until networking (home network and internet providers) gets to a stage where your HDD is the limitation, I can't see a big movement over to SSDs until it becomes cost effective to.
Crappy networks are more a barrier to progress than SSDs/HDDs, reliable 4G/5G networks rolled out for 100% coverage with enough resources put in to give unlimited data for all uses (mobiles/tablet/desktop/home server). Surely this is cheaper than taking a fibre optic cable to each and every house?
Kutark - Thursday, October 22, 2015 - link
4g is only as fast as it is because there are bandwidth caps. People aren't apeshit pounding the network like they do with their land lines. If you had as many users as landlines do, all using it as much as they would normally use their land lines, it would absolutely decimate those networks. And really, you don't need fiber to each house. Coax is more than capable of stupidly fast internet speeds. The issue is getting the cable/telecom companies to not have monopoly access to local municipalities. In all the places in the US where there is actual competition, speeds are great on cable and prices are reasonable. In the places where the useless city council have given monopoly access to the cable infrastructure to one company, its a crap shoot.But seriously, i can get a 250/50 connection on coax right now. Its ungodly expensive, but it can be done. Fiber is NOT necessary.
zodiacfml - Thursday, October 22, 2015 - link
They use fiber now because it is a physical medium that can last for decades. It's capability for bandwidth seems limitless.SunnyNW - Thursday, October 22, 2015 - link
Speeds here in Washington state have been increasing rather quickly in the past year or so. Most of the surrounding state especially Seattle suburbs only have access to Comcast yet comcast has been steadily increasing speeds. At the end of last year, 2014, comcast increased our home speed from 50mps to 105mbps for the same cost per month. After 5 or 6 months of having 105mps I called comcast and asked to be downgraded to 50mbps to save some money. Now again, just this month, they upped our speed from 50 to 75mbps, and I'm fairly certain if I had stayed at 105 that that package was increased to 150mbps. To be honest I don't believe most people even need speeds over 50mbps, I sure don't. It doesnt really affect me much whether my downloading speeds are at 6MB/s or 10MB/s, I still enjoy the same quality of streaming and internet usage, and my household usage on average is about 400GB/month. Also all of the these speed tiers are fairly affordable especially when they give you promotional rates for 6 months to a year and even at the end of the promotion if you call they usually, without much fuss at all, extend your promotional period. Also it only costs an extra $10! dollars per month to get bumped up to the next speed tier, which used to mean going from 50 to 105 in my case a few months ago...I'm not sure what the diff speed tiers are now, now that they have increased their speeds again.My purpose for this post was in reference to all the comments about Google fiber, and gigabit speeds, and such that I always see... I don't think most private residents need anywhere close to gigabit speeds. I will agreee that businesses are a completely diff story and yes there are special cases out there whom I would consider the enthusiasts/power users (including online gamers who want low latency but again they are in the minority). Even with just 75mbps here at home I show about a 9 to 10ms ping consistently, even across large distances the ping is fairly low.
Also one last thing comcast is great at always delivering speeds that are always better than what you would expect. With the 50mbps package I always averaged somewhere in the 60s(mbp/s). Now having 75mbps I avg around 85-90.
In No Way am I pro Comcast but I think internet speeds, especially in my area and I would guess in most urban areas, are to the point where most private residents should be satisfied. Also even without any competition comcast still continues to upgrade their speeds around here. I can easily subscribe to speeds upto 250mbps for about $150/month and in my opinion if you need speeds like that then that price point shouldn't be too far out of reach. You can easily get speeds of 150mbps for less than $100 around here.
I am not naive and do understand that not all areas,especially around the world, have access to decent speeds but I have seen great progress being made in the past couple years, in the US. Would I love gigabit fiber, of course...Do I 'Need' gigabit fiber, if I'm being practical, of course not and I would argue that that would apply to most Americans at least for the immediate future.
Murloc - Thursday, October 22, 2015 - link
well in most places in the world speeds per $ monthly aren't progressing as fast as storage capabilities and file size.For example where I live I get 20 mbit but for like 15 years I've not seen any free upgrade or anything. And I'm in one of the fastest countries in western Europe.
If you take countries like Romania this problem doesn't exist of course. That's just because a country with not even 10 million people has hundreds of ISPs. Other countries are bogged down by the inheritance of state monopolies.
Wolfpup - Thursday, October 22, 2015 - link
This was a surprise, but it makes sense. I'm surprised by how slow the hard drive companies are/were to get in to SSDs!