tealeaf

Does mechanical storage have a future?

Recommended Posts

Something about transfer speed though. My drive now starts at 70-75MB/sec and goes down to about 45MB/sec (according to HDtach). A drive from about 5 years ago did about 15MB/sec, I would guess. So, say flash starts to beat HDD in price/GB in a few years, would it be as fast or faster than HDDs 5 years from now? I have no idea how fast either one will be in 5 years. Just trying to compare flash in 5 years to HDDs in 5 years, since it's useless to compare either one to current technologies. But in 5 years, I would hazard a guess we will be into multiple TB drives. Seagate tells us a 960GB drive will be out before 2006 is over, right? That's nearly a double in about 1.5 years. If HDDs continue to do what they have done for 25 years, they will continue growing at exponential rates per year. Which means we won't have 2TB drives in 5 years, we will have 5TB drives. Unless some huge, unforseen barrier is hit. Perpendicular recording seems to have done away with the limits of horizontal recording. Plus, at least I have no idea what new head or medium technologies will be around in 5 years. There's always room for another late 90s type explosion. I just continue rooting for HDDs since the mechanical complexities of them are so much more interesting, to me, than solid state technologies.

Share this post


Link to post
Share on other sites

It's almost always cheaper to modify existing mass production processes than to develop entirely new ones. That's why there's been so much effort given to developing HDD technlogy over the years. It's likely this trend will continue, but a few companies may be willing to fund alternatives if it looks like they may become cheaper to research and produce than equivalent hard drives in the future.

To me, MRAM seems the most elegant solution, but it's still years away from useful capacities at reasonable prices. In the mean time, flash will do for a growing minority of tasks - pagefile and bulk media storage being the last to go.

Share this post


Link to post
Share on other sites

I also agree with K15. Hard drives in their mechanical state are incredible and they are just so exciting in the way they do their duty reliably 99% of the time with the head flying over the platter at a hair's distance. My belief is that mechanical hard drives do have a long future ahead. Secondly, the industry does not want to produce storage devices that last 30+ years. Hard drives last the useful life of a computer system which is their goal.

They don't make dishwashers to last 50 years because they want people to buy new ones after 10 years. The industry calls for a relatively short lifespan and mechanical drives do that reliably, cost effective and consumer friendly as of yet.

Share this post


Link to post
Share on other sites

Enthusists who don't mind lotsa drives in a system will be replacing their multiples of smaller Raptors and SCSI 15K drives with multiple memory based drives within 3 years. Maybe even within 1 year. These newer drives will nudge out the aforementioned older drives and find use as OS / Programs / Scratch drives and so on.

Just as WD has begrudgingly acknowledged (with Raptor X) that enthusiasts played a significant part in making the Raptor market work, memory based drive makers will get some decent sales numbers out of the enthusiast community. This will seed the expansion of these drives in the broader market until they become commodities.

Already OEM action in the notebook market is looking to play an early role with hybrid and SSD drives as the manufacturers look to cut weight and power consumption while increasing performance. OEMs play(ed) a role in the Raptor arena, although this was perhaps belated hindsight.

Three weeks ago Dell was selling their rebranded 160GB version of the newer 10K Raptor for $180. And so it will be for memory based drives someday. SSDs are a naturally superior technology that was previously hindered in the HDD market only by price. Just think what'll happen when Dell gets em in quantity.

I-RAM was just the beginning and I-RAM 2 is on the way. Flash will make further converts at 32GB and 64GB. For every price drop more will join. And why not? Indications are that Vista will happily live in 20GB or less. OS' are just not expanding in size at the same rate as storage capacities are expanding. And that's a great thing.

Adobe enthusists and pros alike crave a fast scratch drive.

I even believe that within 15 years I might have a chance to put all my data on an SSD. I'm not much into video - music and photos tend to take up the most space on my data drives, so my needs are more modest than some. But still, I think its possible.

Memory based drives might even get large enough in my lifetime to hold my backup data. At which point, for me at least, mechanical storage has no future :P

Edited by hrrmph

Share this post


Link to post
Share on other sites

The only problem is people will see new PCs coming with 30-60GB of space and will wonder what the hell is going on. Almost all new PCs now come with 160GBs now, many times more. We will be into TBs soon. Not to mention, people need LOTS of storage these days. My satellite receiver has a 250GB drive in it, it holds about 25 hours of compressed HD. That's not that much, I would LOVE to have a 1TB drive in there. The firmware doesn't accept it, but the next generation of HD receivers will get 700 and 900GB drives or more. Hard drives are EXCELLENT for markets like that. Performance is a non-issue and they have HUGE, GIGANTIC capacity for the price paid. The market for consumer electronics like PVRs and HD-PVRs will only get bigger. Like I said before, I think in order for solid state technologies to surely, without a doubt overthrow hard drives, they have to be BETTER than hard drives, in every way. Having better performance is only part of it. Like you said, only enthusiasts will buy 32GB of solid state for the same price as a 500GB HDD. 97% of other people won't look twice. The raptor was mostly a niche product, compared to ordinary 7200 IDE drives. New technologies will similarily remain niche products until they can beat HDDs in every facet. Performance, price, capacity. As much as we don't admit it, only a small number of people actually care about hard drive performance. A PC booting in 30 seconds is plenty fast enough for most people. Anyhow, this is just how I see things going down.

Edited by K15

Share this post


Link to post
Share on other sites
Like you said, only enthusiasts will buy 32GB of solid state for the same price as a 500GB HDD. 97% of other people won't look twice. The raptor was mostly a niche product, compared to ordinary 7200 IDE drives. New technologies will similarily remain niche products until they can beat HDDs in every facet. Performance, price, capacity. As much as we don't admit it, only a small number of people actually care about hard drive performance. A PC booting in 30 seconds is plenty fast enough for most people. Anyhow, this is just how I see things going down.

A couple years ago I paid around $200 for a 7200rpm 60GB (Hitachi of course) drive for my laptop. Then I bought another one for my next laptop, and then I bought a 100GB drive when those became available. (Have a spare 60GB 7200rpm drive now, by the way.) Regardless of the $/GB trend for desktop storage, the laptop market is still in a completely different ballpark.

OK, so today I can get an 8GB CF card for $173 at Newegg, ~$21/GB vs the 100GB drive for $160, $1.6/GB. That's already a decent reduction since I posted previously (Jun 8, above, $25/GB). About 16% reduction in barely a month; extrapolating that out 12 months gives us the 8x price reduction in one year, mentioned in my previous post. The 60GB 7200rpm Travelstar was unmatched on the market for a couple of years, and the price remained around the $200 mark for that whole time, so one cannot assume that laptop hard drives will make significant improvements by this time next year. As such, I fully expect that 12 months from now I'll be buying a 128GB SSD at around $400 to replace my 100GB 7200rpm Travelstar.

My desktop system will be another story, but I think the laptop market will completely transition away from mechanical disks within two years.

Share this post


Link to post
Share on other sites

I dont quite agree with your extrapolations, there.

That drop in 4 weeks is just an artefact because of the newly introduced size, and the boost of that last 4 years is also partly because of economy of scale really getting in motion combined with a dropping demand in the cell-phone sector (which had caused a lot of fab expansion).

My guess would be more like 4-5 years for subnotebooks, more for desktop replacement notebooks.

It would be just such a waste to store data like video on flash...

But generally i agree with the fact that small hds will go away sooner.

I really doubt that we will see anything new with 1" or 0.85" platters, because they just dont cut it against flash.

1.8 could have a year or 2 of life in it, though.

But i really cannot see flash replacing the bulk mass storage role 3.5" HD currently inhabitate the next decade.

Stuff like (HD)DVRs lives or dies with $/GB, and even a facter of 2 or 3 would still be signifcant.

Share this post


Link to post
Share on other sites

Like you said, only enthusiasts will buy 32GB of solid state for the same price as a 500GB HDD. 97% of other people won't look twice. The raptor was mostly a niche product, compared to ordinary 7200 IDE drives. New technologies will similarily remain niche products until they can beat HDDs in every facet. Performance, price, capacity. As much as we don't admit it, only a small number of people actually care about hard drive performance. A PC booting in 30 seconds is plenty fast enough for most people. Anyhow, this is just how I see things going down.

A couple years ago I paid around $200 for a 7200rpm 60GB (Hitachi of course) drive for my laptop. Then I bought another one for my next laptop, and then I bought a 100GB drive when those became available. (Have a spare 60GB 7200rpm drive now, by the way.) Regardless of the $/GB trend for desktop storage, the laptop market is still in a completely different ballpark.

OK, so today I can get an 8GB CF card for $173 at Newegg, ~$21/GB vs the 100GB drive for $160, $1.6/GB. That's already a decent reduction since I posted previously (Jun 8, above, $25/GB). About 16% reduction in barely a month; extrapolating that out 12 months gives us the 8x price reduction in one year, mentioned in my previous post. The 60GB 7200rpm Travelstar was unmatched on the market for a couple of years, and the price remained around the $200 mark for that whole time, so one cannot assume that laptop hard drives will make significant improvements by this time next year. As such, I fully expect that 12 months from now I'll be buying a 128GB SSD at around $400 to replace my 100GB 7200rpm Travelstar.

My desktop system will be another story, but I think the laptop market will completely transition away from mechanical disks within two years.

Good points for the laptop market. There, flash does offer some good benefits, in power use, noise, perhaps performance. We are still farther away though. If flash does meet up in price/GB for laptops as soon as you say it will, the market still has to shift, which will take very much longer than that. We are talking about HDD's future and when it comes to laptops, even if flash becomes a good alternative tomorrow, there will be at least another couple years of laptop HDDs. It tends to take a long time for things like this to change-over. Hell, even the SATA change hasn't been clean or quick. People and companies saw no real major reason to switch, so the switch happened when there was no reason NOT to (new system and drives for example) thus the problem with flash overtaking HDDs.

Share this post


Link to post
Share on other sites
Good points for the laptop market. There, flash does offer some good benefits, in power use, noise, perhaps performance. We are still farther away though. If flash does meet up in price/GB for laptops as soon as you say it will, the market still has to shift, which will take very much longer than that. We are talking about HDD's future and when it comes to laptops, even if flash becomes a good alternative tomorrow, there will be at least another couple years of laptop HDDs. It tends to take a long time for things like this to change-over. Hell, even the SATA change hasn't been clean or quick. People and companies saw no real major reason to switch, so the switch happened when there was no reason NOT to (new system and drives for example) thus the problem with flash overtaking HDDs.

I guess 2 years may be too soon for the entire market to change over. But it's not quite the same (to me) as the switch to SATA - that offered pretty much zero benefit to laptop buyers. Regardless of interface speed, the drives are still the same as they ever were. With flash, at least you'll get noise, power, and reliability improvements.

Share this post


Link to post
Share on other sites

Techreport has just tested a 2.5" flash disk, and the results are somewhat surprising: http://www.techreport.com/reviews/2006q3/s...de/index.x?pg=1 Despite a peak burst STR of 14.2 MB/s, it is not always dead last. It actually does pretty well except in STR dependent benchmarks, though note it has a 0.6ms random access time which is excellent for an IDE flash device. Something for those criticizing the "poor" 150MB/s STR of the iRam to think about.

Shockproof, silent, low power consumption, and ok performance. The only big question mark is how many write cycles will it really last?

Share this post


Link to post
Share on other sites
Techreport has just tested a 2.5" flash disk, and the results are somewhat surprising: http://www.techreport.com/reviews/2006q3/s...de/index.x?pg=1 Despite a peak burst STR of 14.2 MB/s, it is not always dead last. It actually does pretty well except in STR dependent benchmarks, though note it has a 0.6ms random access time which is excellent for an IDE flash device. Something for those criticizing the "poor" 150MB/s STR of the iRam to think about.

Shockproof, silent, low power consumption, and ok performance. The only big question mark is how many write cycles will it really last?

How many write cycles did that 160GB Seagate 2.5in perp. rec. drive last? Answer, not many :(.

bfg, go back to my last ? post on Gilbo's SSD Gigabyte thread and read the comments from the link to the akibara(sp?) article, FS and others perpetual urban legends about write cycles. Don't look at last years technology, it's constantly changing (ok, some of the stuff coming out of Asian countries, including India is pretty crappy, but in the long run the quality will get better). 275MB/s for a Viper ram drive. Sure it costs a fortune, but that's a high-end HD camcorder recording device, the camcorder itself is over-priced marketed in limited numbers for TV stations, commercial movie production companies. Look at the low-end Panasonic HD cam, for $5k, and the SSD raid card for that, still limited production numbers, means high prices.

Now look at the long delayed upgrade to the iPod line, supposedly a result of Samsung & others inabililty to supply sufficient numbers of larger capacity low-end performance NAND chips. Given enough time, both price and speed to match or exceed iRam will probably be a reality by the end of next year (hyc, is I believe a little too optimistic, maybe $400 128GB, but not parity with speed/performance of perp. drives of next year). To be sure however, with the market major consumer & business markets going to laptops (leaving Eugene and other high-end IT/network storage freaks in a niche/limited narrowing market overall) over desktops in the next 5 years (Apple already sells more laptops than desktops), you'll necessarily see SSD's in laptops. In the near term. Not price parity, but 5 years out, I'll bet price of SDD's will end up less, and performance including STR's greater than HD's. Reliability, all depends on quality of parts, country of manufacture, quality of manufacturing processes, etc. ; not just the inherent technology itself.

Nothing to get excited about, it will happen, when it happens; I'm not an early adopter unless the price is right, it's only a matter of time, and I cannot predict that timeline any better than Samsung can ;).

Share this post


Link to post
Share on other sites
Techreport has just tested a 2.5" flash disk, and the results are somewhat surprising: http://www.techreport.com/reviews/2006q3/s...de/index.x?pg=1 Despite a peak burst STR of 14.2 MB/s, it is not always dead last. It actually does pretty well except in STR dependent benchmarks, though note it has a 0.6ms random access time which is excellent for an IDE flash device. Something for those criticizing the "poor" 150MB/s STR of the iRam to think about.

Shockproof, silent, low power consumption, and ok performance. The only big question mark is how many write cycles will it really last?

1 million cycles is pretty useful. Remember that this is only a limit on *erase* cycles, and the flash chip's built controller does wear-leveling transparently. I.e., you write a bunch of blocks, the controller writes them to flash. You overwrite those blocks, the controller will just map to a new range of unused blocks. Only when there are no free blocks left will it start erasing unused blocks and rewriting new data onto them.

The one thing I find unforgivable here is the pitiful capacity for the price. Considering that 8GB of flash fits completely in an SD card, 99.9% of that drive is dead weight. These guys aren't even *trying* to offer good value, and for that stupidity this product deserves to fail.

Share this post


Link to post
Share on other sites

It wasn't so very long ago that everyone thought that a 4.7 GB hard drive was more than one would ever need, wasn't it?

I don't think anyone ever thought 4.7GB hard drive is enough, but someone did think 640K of memory is enough for anybody :huh:

Really who? This statement is commonly attributed to Bill Gates but there is no evidence he ever said it and he denies it http://en.wikiquote.org/wiki/Bill Gates#Wrongly Attributed. It's possible someone at IBM said/thought it but I doubt it. It was simply a design decision based on the expected needs etc at the time and in the immediate future (good or bad one that's up to you to decide).

Just read this now, worse than that was the lack of drive barriers being considered, by many companies (microsoft most certainly included). Until FAT32, microsoft never kept more than 3 years ahead of hard drive developement. Same with BIOS and motherboard companies. Really, if they had to design a new file system/bios/adressing routines, they may as well have went nuts and, in 1990, said "this interface will handle 100GB of addressable space". Can't really blame them though, humans don't naturally think in exponential numbers. Although, PC hard drives were only 9 years old and no one could forsee the changes to happen in the 1990s. 100MB to 100GB in 10 friggin years.

I even remember getting a new PC in 1999 with a small (even by then) 8GB Maxtor drive. It replaced a 1.2GB western digital and I thought "holy stinker, how can I fill 8GB of hard drive?!!!" Within one year, it was close to being full of MP3s.

Share this post


Link to post
Share on other sites

It wasn't so very long ago that everyone thought that a 4.7 GB hard drive was more than one would ever need, wasn't it?

I don't think anyone ever thought 4.7GB hard drive is enough, but someone did think 640K of memory is enough for anybody :huh:

Really who? This statement is commonly attributed to Bill Gates but there is no evidence he ever said it and he denies it http://en.wikiquote.org/wiki/Bill Gates#Wrongly Attributed. It's possible someone at IBM said/thought it but I doubt it. It was simply a design decision based on the expected needs etc at the time and in the immediate future (good or bad one that's up to you to decide).

Just read this now, worse than that was the lack of drive barriers being considered, by many companies (microsoft most certainly included). Until FAT32, microsoft never kept more than 3 years ahead of hard drive developement. Same with BIOS and motherboard companies. Really, if they had to design a new file system/bios/adressing routines, they may as well have went nuts and, in 1990, said "this interface will handle 100GB of addressable space". Can't really blame them though, humans don't naturally think in exponential numbers. Although, PC hard drives were only 9 years old and no one could forsee the changes to happen in the 1990s. 100MB to 100GB in 10 friggin years.

I even remember getting a new PC in 1999 with a small (even by then) 8GB Maxtor drive. It replaced a 1.2GB western digital and I thought "holy stinker, how can I fill 8GB of hard drive?!!!" Within one year, it was close to being full of MP3s.

Share this post


Link to post
Share on other sites

It wasn't so very long ago that everyone thought that a 4.7 GB hard drive was more than one would ever need, wasn't it?

I don't think anyone ever thought 4.7GB hard drive is enough, but someone did think 640K of memory is enough for anybody :huh:

Really who? This statement is commonly attributed to Bill Gates but there is no evidence he ever said it and he denies it http://en.wikiquote.org/wiki/Bill Gates#Wrongly Attributed. It's possible someone at IBM said/thought it but I doubt it. It was simply a design decision based on the expected needs etc at the time and in the immediate future (good or bad one that's up to you to decide).

Just read this now, worse than that was the lack of drive barriers being considered, by many companies (microsoft most certainly included). Until FAT32, microsoft never kept more than 3 years ahead of hard drive developement. Same with BIOS and motherboard companies. Really, if they had to design a new file system/bios/adressing routines, they may as well have went nuts and, in 1990, said "this interface will handle 100GB of addressable space". Can't really blame them though, humans don't naturally think in exponential numbers. Although, PC hard drives were only 9 years old and no one could forsee the changes to happen in the 1990s. 100MB to 100GB in 10 friggin years.

I even remember getting a new PC in 1999 with a small (even by then) 8GB Maxtor drive. It replaced a 1.2GB western digital and I thought "holy stinker, how can I fill 8GB of hard drive?!!!" Within one year, it was close to being full of MP3s.

Share this post


Link to post
Share on other sites
Guest 888

Theoretically, it's possible to produce about 250TB capacity in 3,5" HDD format before the last physical limit is reached. It is realized on SOMA media, in combination with HAMR which could offer an ultimate areal density of 50 Tbpsi. Of course it may be still too far away from now but... just using HAMR (Heat Assisted Magnetic Recording), it's very realistic to get 20...50TB 3,5" HDDs. Still, also this HAMR technology needs to wait for some more years to reach into mass production but it is researched deeply by Seagate already from 2000 or so, so if the perpendicular technology reaches its limits (about 3...6TB in 3,5" HDD) one time, then I think the HAMR is ready to start...

So I think, the mechanical storage has still long future from now, at least in big storage-HDD segments.

Share this post


Link to post
Share on other sites

http://www.realtechnews.com/posts/3258

New Protein-Based Storage Technology Will Someday Boost DVD Capacity to 50TB

Further proof the Blu-ray and other DVD technologies are just a quick stop on the road to even higher density DVD storage, an Indian scientist today announced a breakthrough that could someday boost DVD storage to a whopping 50 terabytes and make the hard drive obsolete. By coating the disc surface with a light sensitive protein, the capacity boost from the small genetically altered microbe proteins could eventually make a disc surface capable of holding 50,000 gigabytes

Professor V Renugopalakrishnan of the Harvard Medical School in Boston has claimed to have developed a layer of protein made from tiny genetically altered microbe proteins which could store enough data to make computer hard disks almost obsolete. “What this will do eventually is eliminate the need for hard drive memory completely,†ABC quoted Prof. Renugopalakrishnan, a BSc in Chemistry from Madras University and PhD in biophysics from Columbia/State University of New York, Buffalo, New York as saying.

The new protein-based DVD will have advantages over current optical storage devices such as the Blue-ray as well, because the information is stored in proteins that are only a few nanometres across. “The protein-based DVDs will be able to store at least 20 times more than the Blue-ray and eventually even up to 50,000 gigabytes (about 50 terabytes) of information. You can pack literally thousands and thousands of those proteins on a media like a DVD, a CD or a film or whatever,†he said. Source: Yahoo News

We Say: Only downside I see is the inevitable ultra-extended director’s cut. King Kong was just not long enough! And what on earth would a single disc cost? $1,000? not to mention the player. Sounds interesting, but don’t hold your breath on seeing this anytime soon.

Share this post


Link to post
Share on other sites

When I first saw that the other day I thought "Great, now my data can rot away." ;)

As for the STR issue, Sandisk has had flash cards doing "Enhanced Super-Parallel" internal RAID-0 for what, two years now? Considering how much more reliable flash is than mechanical (at least at first) going massively RAID-0 internally would give you any STR you could want up to the interface limits. Unfortunately Sandisk chose to make their Extreme III compactflash cards do PIO modes 5 and 6 rather than any DMA mode, or we'd all be using them already with ATA to compactflash adapters. The 2.5" Super Talent disk that Techreport tested does use DMA.

Since RAID-0 also solves the capacity issue, the main barrier to adoption is bringing the price point down.

Share this post


Link to post
Share on other sites

I would be careful to say "how much more reliable" flash is than mechanical. It IS more reliable, but when we are talking about less than 1% failure rates, it's hard to be "much more" reliable. Most of the solid state parts of a computer have a much higher failure rate than a hard drive, FYI.

Share this post


Link to post
Share on other sites

Comments about solid state reliability.

I've read that about half of HDD failure rate cannot be foreseen by using SMART. SMART can predict mechanical failure, so about 50% of HDD malfunctions are related to solid state components on the PCB of HDDs. Why else do people swap PCBs to recover data? Is solid state really that more reliable after all? More shock tolerant they may be, though. And solid state memory differs from solid state spindle motor controllers and other high power-usage IC chips.

IMHO, solid state is very much welcome into laptops. Not much reason to have a HDD constantly spin up and down draining the batteries, etc. Also solid state can eliminate seek time in some server environment where small capacity is of no issue.

______________

Comments about protein based DVDs.

Mad Scientist: [/i]“What this will do eventually is eliminate the need for hard drive memory completelyâ€

Maybe... that is IF all the following is correct:

- Protein DVD appears soon enough as HDD capacity increases rapidly as well.

- Protein DVDs are cheap enough.

- Protein DVD drives are cheap enough. (One would require at least two DVD drives if one intends to replace the system HDD with a DVD.)

- Protein DVDs would have to be DVD-RWs with unlimited (or nearly unlimited) rewrite cycles.

- Protein DVDs should be able to withstand UV-radiation (i.e sunlight) without damage or data loss, ambient heat without protein breakdown, in addition to be immune to fingerprints and dirt.

I find it much easier to handle HDDs cautiously without banging them against hard objects, than to keep a CD/DVD clean from any slightest amount of dirt or fingerprints. HDDs are pretty much immune to dirt, as HDDs have filters and platters are never to be removed. Would a sealed DVD player with a non-removable single DVD-platter be useful? Probably not.

So if HDDs are to be replaced, at least in some portion (like system files), I'd trust it to solid state memory. DVDs are just too easily damaged (mainly because they are removable media). They are too noisy as well (because the drives are fast but the disks are poorly balanced, unlike HDD platters). Why do they revolve so fast? Are they really trying to compete with HDDs? That's quite similar to Don Quijote's fight agains windmills.

BTW, I use Nero Drivespeed to limit my CD-drive to 15x (CD-drive is capable of 48x). It does silence the CD-drive when it's reading. The problem is, the rpm is reduced after I read some data out of the CD. When I enter a new CD, it will spin at maximum rpm for several seconds. I wish there was some tweak to force my drive to 15x right from the second I insert a new disc...

Share this post


Link to post
Share on other sites

Well, I was just thinking most DOA failures of mechanical drives are probably from improper handling during shipping or installation, and flash would have the advantage there.

The Super Talent drive had no DRAM cache, while Hitachi looks set to add one: http://uk.theinquirer.net/?article=33030

MRAM is now finally available but only in 512kB chips at the moment: http://media.freescale.com/phoenix.zhtml?c...0030&highlight=

The old CD-Rom drives that used caddies definitely protected the discs from dust and fingerprints better. And I've limited optical drive speeds before with Plextools or custom firmware.

Share this post


Link to post
Share on other sites
I dont quite agree with your extrapolations, there.

That drop in 4 weeks is just an artefact because of the newly introduced size, and the boost of that last 4 years is also partly because of economy of scale really getting in motion combined with a dropping demand in the cell-phone sector (which had caused a lot of fab expansion).

My guess would be more like 4-5 years for subnotebooks, more for desktop replacement notebooks.

It would be just such a waste to store data like video on flash...

But generally i agree with the fact that small hds will go away sooner.

I really doubt that we will see anything new with 1" or 0.85" platters, because they just dont cut it against flash.

1.8 could have a year or 2 of life in it, though.

But i really cannot see flash replacing the bulk mass storage role 3.5" HD currently inhabitate the next decade.

Stuff like (HD)DVRs lives or dies with $/GB, and even a facter of 2 or 3 would still be signifcant.

+1

Namely in places like iPods and ultra-portable laptops. Where small slow drives are costing an arm and a leg right now. Once SSD's can reach the same capacity as those small drives I'm sure they'll quickly become the new standard.

Share this post


Link to post
Share on other sites
Comments about solid state reliability.

I've read that about half of HDD failure rate cannot be foreseen by using SMART. SMART can predict mechanical failure, so about 50% of HDD malfunctions are related to solid state components on the PCB of HDDs. Why else do people swap PCBs to recover data? Is solid state really that more reliable after all? More shock tolerant they may be, though. And solid state memory differs from solid state spindle motor controllers and other high power-usage IC chips.

IMHO, solid state is very much welcome into laptops. Not much reason to have a HDD constantly spin up and down draining the batteries, etc. Also solid state can eliminate seek time in some server environment where small capacity is of no issue.

SMART can't predict HDD failure well because many failures (even mechanical failures) are sudden and catastrophic. SMART only helps with gradual failures, such as slow degradation of a magnetic head.

You do have a point, though. Funny that my DiamondMax D540X is still alive and kicking after 5 years, and I had to replace my Athlon CPU due to instability.

Share this post


Link to post
Share on other sites
Now look at the long delayed upgrade to the iPod line, supposedly a result of Samsung & others inabililty to supply sufficient numbers of larger capacity low-end performance NAND chips. Given enough time, both price and speed to match or exceed iRam will probably be a reality by the end of next year (hyc, is I believe a little too optimistic, maybe $400 128GB, but not parity with speed/performance of perp. drives of next year). To be sure however, with the market major consumer & business markets going to laptops (leaving Eugene and other high-end IT/network storage freaks in a niche/limited narrowing market overall) over desktops in the next 5 years (Apple already sells more laptops than desktops), you'll necessarily see SSD's in laptops. In the near term. Not price parity, but 5 years out, I'll bet price of SDD's will end up less, and performance including STR's greater than HD's. Reliability, all depends on quality of parts, country of manufacture, quality of manufacturing processes, etc. ; not just the inherent technology itself.

Nothing to get excited about, it will happen, when it happens; I'm not an early adopter unless the price is right, it's only a matter of time, and I cannot predict that timeline any better than Samsung can ;).

Given 40MB/sec 8GB CF cards today, http://www.dailytech.com/article.aspx?newsid=3419 I'd say we're well on our way. 8GB @ $640 today, 64GB at that price a year from now, so ok, slightly more than 12 months for my $400 128GB wishlist, but I still think less than 24.

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now