Slashdot Mirror
Is SSD Density About To Hit a Wall?
Zombie Puggle writes "Enterprise Storage Forum has an article contending that solid state disks will stay stuck at 20-25nm unless the materials and techniques used to design Flash drives changes, and soon. 'Anything smaller and the data protection and data corruption issues become so great that either the performance is abysmal, the data retention period doesn't meet JEDEC standards, or the cost increases. Though engineers are working on performance and density improvements via new technologies (they're also trying to drive costs down), these are fairly new techniques and are not likely to make it into devices for a while."
Memristor technology doesn't even work with feature sizes that big, so it's the logical next step. Also it can be layered and so leverage Dimension Z. Products expected in three years from a joint HP and Hynix venture. No worries.
Help stamp out iliturcy.
Improving upon current SSDs will require new technology! Isn't that sort of implied in the whole concept of, you know, progress?
The wall or plateu or whatever you prefer to call it of electronics progress is similar to the recurring doomsday predictions. It's always right around the corner, but it never happens.
I guess we could liken it to fusion, strong AI, the second coming of Jesus and whatever else that generally is put in the belive it when see it folder.
It doesn't seem a big deal to me. I'd be more interested in seeing the prices drop and to have larger RAM caches.
At what price learning? At what cost wisdom? The price is a man's peace of mind, and the cost is his life.
Well, the density is already not bad, so the big key is to get the cost down! For larger applications of Flash memory(like over 250GB) I don't think the physical size is going to be a problem because it is competing with 3.5" and 2.5" hard drives.
Aside from cost, there are plenty of other non-density things to work on: number of rewrite cycles, speed, reliability, etc. I can't wait for the day that spinning media eventually goes bye-bye.
The only way a density wall will matter is if price can't scale. If price scaling is purely based on density, then we have a problem as SSDs aren't price competitive yet. However if price can keep scaling down, then no problem. As it stands you can pack in flash to easily meet the same density as magnetic storage. They have 512MB SSDs that are 2.5" form factor and I suspect it could be smaller, it is just that size to fit in regular laptop drive slots.
So we'll see. I don't know what are the price barriers for SSDs, and I'm sure smaller lithography helps, but I suspect they'll be fine without it.
There are far better technologies waiting to replace it, one being P-RAM. The best thing is, none of the newer tech is subject to Flash's crippling block-erase semantics, and so they are far more suitable for SSDs. No longer will SSDs require tremendously complex controllers and firmware in order to attain good performance, allowing new SSDs to be both cheaper, faster, and more reliable.
Can you imagine a stack of 128 8GB micro-SD cards next to a 1TB 3.5" hard disk? Do we need higher densities? Of course there's a relation of price to die area, but there's still room for improvement in manufacturing technologies. As long as the price keeps going down, I'm fine with the current density. Worry about metrics that matter.
"Enterprise Storage Forum has an article contending that solid state disks will stay stuck at 20-25nm unless the materials and techniques used to design Flash drives changes"
Uh huh. And how is this situation different than every single computer technology we've had since 1970?
..times up.
Guess we will be stuck where we are on this matter forever.
Or does this mean there is something that is going to replace this tech anyway?
HP and Hynix are doing memristors, while the entire rest of the industry is doing phase-change memory.
Densities are fine. The main problem is lowering the cost. They need to drop the price by an order of magnitude. I am sure it costs way less than that to manufacture .. they just have to pay back all the research and equipment capital costs and build more production lines. Once they do that it will be dirt cheap. I remember when LCD monitors were a couple thousand bucks. And hard drives were far more expensive than SSDs are today .. and that was only 15 years ago.
For example an OCZ Technology 250 GB SSD is $450 .. I paid around $400 for a 400 Megabyte drive in 1995. That's works out to hard disks back then being nearly 5 times the price per megabyte of SSD drives today.
but who says the wall is going to win that collision? I've seen it time and time again: a problem is encountered, and dealt with. Optical disk rotation speed. Parallel data buses. Processor clock speeds. They all hit a wall, and we got around that wall. We lowered the wavelength of the laser instead of go to 56x CDs. We switched to serial buses when parallel encountered clocking issues. We switched to multicore processors when we couldn't keep upping the gigahertz. I'm fully confident we'll figure out a solution to this problem as well, whether it be new manufacturing techniques, memristors, or just larger Flash chips.
You can get a 32GB MicroSD card, which has a volume of of (15 mm) * (11 mm) * (1 mm) = 0.165cc, yielding a storage/volume of 32GB/0.165cc ~= 194 GB/cc. A 3TB HDD has (roughly) a volume of 102 mm * 25.4 mm * 3.5 in ~= 230cc. This yields a density of 3TB/230cc ~= 13 GB/cc.
Looks like SSD is roughly 15 times denser than a conventional 3.5" drive. Seems to me it's SSD's scalability that needs work -- not the physical density.
Local storage is a lot cheaper and faster for most people in the USA, which is all I can speak of. Maybe over in Utopialand where everyone has 100 gig speed connections and hosting is pennies a day for terrabytes the "cloud" might be cheaper and better. Our domestic broadband speeds and prices are not even close to keeping up with increased local storage density and lowering prices for same. Saying the "cloud" will do everything is sorta naive, we have all the major ISPs talking about limits and caps now. This is 100% the WRONG time to be shifting to far away "cloud" storage for most people.
I know I'll be keeping my movies and files handy right here, thanks. I just can't see storing multiple gig sized movies way over there someplace when it would cost me two cents to store it here and have it playback at fast streaming speeds for the cost of the electricity.
Having to go pay yet again to watch your movie or access your own file..nope. The "cloud" is a marketing buzzword for companies that want to charge you serious coin for access to *your own files*.
You know, stories like this used to interest me. Then I noticed that:
a) they kept reoccurring, and
b) had a common theme.
Yeah, it's always "We're approaching a wall with what can be done with current technology, so it's going to either be more expensive, or need a new technique, yadda yadda." Tell you what. Lemme know when we *actually* hit the wall in ANY of these that they keep threatening us with a wall in making, SSD, HDD, CPU size, etc.
Canada: The US's more awesome sibling.
The smaller the NAND flash process size the shorter the write endurance and data retention times. A 25nm NAND flash SSD will have a much shorter lifespan and hold data for a much shorter period of time than current 34nm tech. Does this mean that 2010 NAND flash SSDs will be better than 2011 ones? Well I guess that depends on how much you value reliability and longevity in your storage devices. Lower cost and shorter life is a win/win for the manufacturers. This limit on NAND flash technology has been known since the start. I don't see the big deal. Just stop at 34nm and work at other technologies that are faster or scale in size better. We usually think of larger process size as being better, but in this case it's not.
http://features.techworld.com/storage/3212075/is-nand-flash-about-to-hit-a-dead-end/?intcmp=ft-hm-m
http://hardforum.com/showthread.php?t=1492711
Quite an experience to live in fear, isn't it? That's what it is to be a slave.
Think of all the desktop cases with space. Layers of chips high and the depth too.
If not, always U2.
Domestic spying is now "Benign Information Gathering"
Got to keep cartel pricing up, why race to the bottom just yet :)
Domestic spying is now "Benign Information Gathering"
I don't think so. Back when I used to do research on microelectronic fabrication methods, we bought 3-inch wafers for about $10 apiece. Those were high purity with doping to whatever type and level we selected. And that was without bulk pricing or favorable price scaling with larger wafers.
Our molecular beam growth chamber, however, cost hundreds of thousands of dollars plus tens of thousands per year for supplies and maintenance (plus tens of thousands for a postdoc and a grad student to run it).
So I really think the cost of equipment and processing far outweighs the cost of the silicon wafers. Otherwise, all CPU's with the same physical size would have roughly the same price, regardless of transistor count or clock speed.
194 GB/cc is about 8e10 atoms per bit, assuming 2 Angstrom atoms. Since it's going to be really difficult to store more than 1 bit per atom, that sets a hard limit of improvement at 8e10 times what's available today.
Contribute to civilization: ari.aynrand.org/donate
The sky is always falling. We never know how to make it to the next generation. If we knew that already it would have already ceased being the next generation.... SSDs are now more or less completely tied to the same fate as CPU's and as such if this were really that big of a problem.. The article would be "Is computing power about to hit a wall"
640PB ought to be enough for anybody.
HTML5 drives are the future.
"We live in a global world" - Harvey Pitt, former Securities and Exchange Commission Chairman
Thought that I just read that almost everyone is *bailing* on PCM and looking into memristors because they can't get the power down on PCM. PCM work is farther along, but many feel that it's not going to happen and feel that memristors have more promise. I, for one, welcome our memristory overlords.
The effects of EM fields can be significantly reduced by conductor pairing. When two currents of equal and opposite magnitude run side by side, the EM field is almost entirely confined to a space around those conductors. This can be achieved by creating cell pairs arranged so they are side by side, but turned in opposite directions. This allows the current of one to be in the opposite physical direction of the other, when the same operation is being performed on each. Since erase and read (but not write) can always be done at the same time, this reduces the number (in the case of read) and severity (in the case of erase) of EM fields, reducing the overall effect of EM fields on adjacent inactive cells.
now we need to go OSS in diesel cars
Density isn't really the story here. HD technology will always have higher density, at least for the next 20 years or so until someone comes up with a cheap way to produce nanoscale chips without using expensive lithography processes.
The real story here is performance. SSDs can scale performance far better than HDDs can. We may be hitting a wall on density but there is plenty of room to boost chip interconnect speeds and make the performance of a SSD scale with the number of individual chips making up the storage. We already have this to some degree. It is going to get a lot better. 6GBit SATA is going to be maxed out in less than a year.
The performance equation is going to radically change how systems scale. Right now computer hardware, sans the storage, tends to be limited by available RAM (compute power is no longer the defining issue in a modern system). As SSDs become faster a new mid-level cache tier becomes more and more viable, so instead of having to stuff the machine full of expensive ram one can instead just put a moderate amount of a ram in (say, 8-16G) and stuff a SSD as a 80G mid-level caching tier.
-Matt
Storing more than 1 bit per atom shouldn't be that hard, really. Atoms have tons of degrees of freedom - discrete excitation levels, spin, charge, the list goes on. Getting a hardware platform to utilize those degrees is a whole 'nother challenge, but saying 1 bit per atom is the hard wall on theoretical grounds is just silly.
The density of a flash based storage device depends on the density on the chip and the density of chips.
SSDs* don't seem to get anywhere near the overall density that the small flash cards do. The largest SSDs in the 2.5 inch form factor seem to be 512GB and the largest SSDs in the 3.5 inch form factor seem to be 1TB. I dunno for sure why this is but I suspect it's because noone is prepared to put the engineering effort required to cram all those chips into a product that at current flash prices would be extremely niche.
Density on the chips is important for two reasons, firstly it increases the ammount of flash you can get in with a given degree of cramming effort. Secondly and more importantly it will drive down the cost.
*defined as flash storage units that can replace and outperform a hard drive.
note: i'm known as plugwash most places but i screwd up registering that here somehow in the past and now can't register
Until 2008 the memristor was a theoretical construct - a presumed fourth element to complete the symmetry between resistor, inductor and capacitor. But then in a moment it went from theoretical to provably found, and the theory became real. It turns out that it took researchers this long to find it because the effect doesn't work at all in larger process sizes. They needed to try it at the recently evolved process sizes to definitively find the effect. Now they have found it, and it works. Since it's a new discovery limited by feature size at 50nm maximum, one would presume that we will need to explore new finer lithography technologies for some time before its minimum feature size is found.
The innate nature of the technology is that it's stackable. It can exploit dimension z. That's not even debatable - it's even given in the fine article. It doesn't rely on dopants embedded in the silicon, but on the junctions between mettalic elements laid upon it. It is fast. Cells are analog so it's possible to store multiple bits in a cell to the limit of how finely the programming current can be regulated, which is a factor that improves over time. It's low-power, and obviously so low-heat. There are some thermal implications for filesystems based on this storage that can best distribute the thermal load of writing, but that's a programming issue easy to overcome.
It's also already small. It doesn't even work on feature sizes larger than 50nm. We won't know how small a feature size it works on until we develop new methods of lithography that work at finer levels of resolution than it works at. It could be quite some time before that happens. We're stretching the limits of ultraviolet already and up from here is X-Rays and Gamma rays, which are hard to produce.
Between the three dimensional elements, the fine resolution elements, the multiple bits per cell elements, the high speed of access and programming, this does look like the technology to carry us forward from flash memory if it can be produced commercially. The partnership between HP and Hynix to implement commercial production does imply that it's coming. They've announced a plan and a schedule. One would presume their engineers are hard at work and the remaining practical questions involve layout of the memory grids to optimize performance to the interface and provide sufficient indirection to deal with inherent physical media reliability.
Help stamp out iliturcy.