SSD Prices On Parity With High-End HDD By 2011

kgagne writes "EMC executives were heavily pitching the virtues of solid state disk drives at their annual users conference in Las Vegas, saying that SSD will not only be on price parity with high-end Fibre Channel disk drives by the end of 2010 or early 2011, but that NAND memory will solve all sorts of read/write issues created by spinning disk technology. EMC's CEO and its storage platforms chief said the company will do everything it can to drive SSD prices down, and adoption up, by deploying them in their products. One issue might be that EMC is using SSD from STEC, which is being sued by Seagate for patent infringement." The article also mentions some of the work EMC has been doing to make sure SSD is enterprise-class reliable, such as developing "wear leveling" software.

Comments

SSD from STEC

[quarrel]

> One issue might be that EMC is using SSD from STEC, which is being sued by Seagate for patent infringement.

Why is this an issue? If EMC think the technology is a winner, and they don't have a stake in a particular player (of course they have to choose a supplier, but that hardly indicates a long term commitment) then what do they care who wins?

One of the great things about being in EMCs shoes is that you want these things commoditized.

Either way, as a the sooner SSD is directly competitive the better. They're ICs - you churn them out, and only worry about yield. HDDs are mechanical and will always have their mechanical shortcomings.

--Q

Re:SSD from STEC

[vax]

if seagate has the patent then they better start making some damn SSD drives. (that are actually on the market)

it would be good to have some competition anyway to drive the prices down..

still. I cant wait for 100gb SSD drives.. finally a laptop for gigging that can handle a beating.

really once these are standard in laptops I think you will see more robust laptops on the market since the spinning disks have always been one of the quickest parts to fail (well assuming that the laptop has decent cooling design and isnt running some desktop processor that is melting the damn thing)

The Future is Solid State

[Bananatree3]

Spinning disk hard drives are the mode, median and mean today. You can grab a 1TB platter hard drive for under 200 bucks. It may not last as long as a SSD, but at that price you can certainly buy a bunch of backup drives for a lot cheaper than a 1TB Solid State drive.

However, SSD is the future wave, as it Just Works better than platter drives. A high quality, high density, low priced SSD would knock the socks off any platter drives today if it were available. Platter drives will be the mainstream market for a while because of cost and size availability. However as SSDs become cheaper and hold more space, the WILL push platter drives out.

Re:The Future is Solid State

[arbiter1]

i've had platter hard drives go 8+ years, and one i know of is at 9 years old and still goin strong

Re:The Future is Solid State

[hostyle]

-1 Re-Iterating The Damned Obvious

Re:The Future is Solid State

[kesuki]

"One cannot know whether or not the SSD makers "lie the same" as the disk makers."

Sure you can, they're Human of course they lie. MTBF can be generated, based on how long it takes for 'more than 50% of the data sectors fail' but who would keep using a disc that kept having randomly failing disc sectors, even if SMART technology can reduce the risk of loosing data...

so of course SSD makers are going to calculate a MTBF assuming the type of wear a typical person who boots vista once, for every time it crashes*, and does nothing but plays solitaire, freecell and spider solitaire all day! it takes a lot of hours of playing solitaire, a non disk heavy activity for a SSD device to fail!

remember NAND flash memory is based on changing the structure of a semi-conductor with an electric charge, the more you do it, the sooner the device fails... even reading the state of the material causes wear, because electricity is used to read as well as 'write' to the memory, kinda the way a laser erodes the dye on a recordable optical disc... obviously though if you can Write millions of times, you can read billions of times. there is another problem with NAND though, NAND memory is often made with tantalite, a rare mineral used in semiconductors and capacitors..

if we recycled 100% of computer parts, the tantalite problem would be solved easily, but we're not even close to 10% recycling... so even as we speak gorilla habitat in Africa is being destroyed for tantalite mines. they've been building more and more of the mines, since a price spike in 2000, and the number of mines running are keeping the cost of tantalite from spiking again, but it would be so much better if we just recycled our old computer waste. we could save gorillas, if we pushed for refundable deposits on recycling electronic devices like computers, etc. if it worked for the lead acid battery it can work for tantalite, copper, aluminum, gold, and silver in electronic devices. FWIW plastic in electronics can be recycled into diesel fuel, as i found out from here. http://www.cbc.ca/news/background/small-business/recycling.html

*= which is probably every day, more if you install hardware drivers.

Re:The Future is Solid State

[renoX]

The future is very near!
Sure currently buying a 1TB Solid State drive would be too expensive, but do we need really it?

No: on my HDD, I have two partitions: one of 30 GB for the OS and the software (which has still a lot of free space), and a big one for the data.

Replacing the OS&software partition with a SSD would bring 99% percent of the benefits of having a 'full' SSD: fast boot time, fast application startup, etc.. Especially as we can use a part of the SSD as a cache for the HDD.
So IMHO, we don't really need big SSDs (30-60GB is enough), but fast SSDs should improve the computers' responsivenesscomputers..

So my ideal computer would have: one fast SSD of 60GB (for responsiveness), 1TB of HDD (for bulk storange) plus another external 1TB HDD for backups.

What about filesystems...

[epiphani]

Given that many filesystems are designed specifically with the spinning magnetic disk in mind, what open source filesystems are out there that will work to the advantages of solid state storage? Has anyone started thinking about that one as something to address before the major switches start taking place?

Or... does solid state storage take care of those oddities in firmware with the whole automatic write leveling technology?

Re:What about filesystems...

[v1]

Driver level can make certain assumptions about the physical drive, such as seek time, and for example, work to decrease disk fragmentation. Fragmentation is very minor issue with SSDs. So there will be a minor performance hit (from maximum possible in the SDD) due to the things the drivers and os do to try to get the most performance out of a HDD.

The only adaptation I can see is trying to minimize wearing on certain blocks, but from the looks of it the SDD's are being designed with wear leveling in mind so I doubt even that will matter to the software.

I could see other minor tweaks. I'm sure no OS seriously expects a new hard drive to spin up reliably in anything under 2 seconds. Imagine how fast wake times on laptops can be when restoring RAM from storage? As long as the hardware is being worked on to wake up that fast. But right now they know they have a few seconds to wait for the HDD to spin up so they're not necessarily seeing a need to optimize wakeup. I'm sure there are other similar issues.

A very useful change would be to alter the standard block size from 512 bytes to something larger, say 32k. Since it's more efficient to flash larger blocks at a time, we may see native block sizes go up for optimal performance on SDDs. No telling how well the OS and software will handle that sort of change. I bet that is hardcoded all over the place. That would dramatically improve write speeds though. So there will be growing pains

Re:What about filesystems...

[tooyoung]

Given that many filesystems are designed specifically with the spinning magnetic disk in mind, what open source filesystems are out there that will work to the advantages of solid state storage? Has anyone started thinking about that one as something to address before the major switches start taking place?

No, no one has even considered that yet. I'll alert the academic world while you clue in the industry.

Re:What about filesystems...

[4e617474]

Flash drives have been around for a while, you know. And so have the filesystems:

YAFFS and JFFS2 look to me like they might be showing their age.

From Wikipedia:

"YAFFS2 is similar in concept to YAFFS1, and shares much the same code... The main difference is that YAFFS2 needs to jump through significant hoops to meet the "write once" requirement of modern NAND flash.

YAFFS2 now supports "checkpointing" which bypasses normal mount scanning, allowing very fast mount times. Mileage will vary, but mount times of c. 3 seconds for 2 GB have been reported.

Measuring mount times in seconds per gigabyte is not encouraging for the design goals we're talking about here. The disadvantages section of the JFFS2 article pretty well speaks for itself, but note

"All nodes must still be scanned at mount time."

Overcoming that hurdle was how YAFFS2 even moved up to the seconds per gigabyte range - the introductory paper for LogFS says

"On the authors notebook, mounting an empty JFFS2 on a 1GiB USB stick takes around 15 minutes. That is a little slower than most users would expect a ïlesystem mount to happen."

The developer's gift for dramatic understatement aside, LogFS sounds like they intend to meet the challenge of what's actually next head-on but the home page still has nice tidbits like -

"http://wiki.laptop.org/go/Logfs - this advertises several non-existent features, so don't take it too seriously. On the other hand, most of them will get implemented over time."

Note that the link above only lists six features so for "several" to be non-existent, well... So no, it's not a given that the problem of a workable filesystem is long since solved, or that a suitable one will actually be ready for prime time when they've got the SSD hardware at the "sweet spot".

Re:What about filesystems...

[kesuki]

The only adaptation I can see is trying to minimize wearing on certain blocks, but from the looks of it the SDD's are being designed with wear leveling in mind so I doubt even that will matter to the software. Actually with proper software you'd probably like to do the opposite - try to wear out certain blocks as fast as possible. This way the lossage is more predictable and rest of the disk is kept in a better shape. Point being that bad sectors aren't really a big deal if you're prepared for those. When NAND memory fails, it can fail in such a way as to make the ENTIRE flash memory device unreadable... this is from real world NAND memory devices failing from real world use, all of a sudden not wear leveling seems like a suicidal mode of wear... if the entire chip can short out from a single block failing.

"In case of a massive damage,

        * If the device is not accessible at all (circuitry failure), no software can even attempt the recovery. Physical intervention is required.
        * Even if the device seems accessible, software recovery run the will take excessive time to complete, making the attempt impractical. On top of that, the recovery run puts further stress on the device. This may be undesirable.

In case of the massive damage, there is no point in attempting the do-it-yourself type data recovery at home. There is little you can do to repair a physically damaged device without the special equipment. If you have a physically failed storage device, we have a discount available for a DriveSavers recovery service. DriveSavers are quite good with physically damaged devices and we recommend you contact them if need arises."

http://www.z-a-recovery.com/physical-flash-memory-failure.htm

Re:What about filesystems...

[kesuki]

When NAND memory fails, it can fail in such a way as to make the ENTIRE flash memory device unreadable... this is from real world NAND memory devices failing from real world use, all of a sudden not wear leveling seems like a suicidal mode of wear... if the entire chip can short out from a single block failing. There's 0 reason for a properly designed flash device to fail like that due to wear, leveling or not. That's just shitty engineering. "Tunnel injection is the quantum tunneling effect, also called Fowler-Nordheim tunnel injection, when charge carriers are injected to an electric conductor through a thin layer of an electric insulator."

you should have said 'IANAEE' for i am not an electrical engineer. because the way NAND ram works it is entirely possible for failure to be a complete and total failure of the device, or at least 512 blocks of the device, if it doesn't create a short that prevents the whole device from working.

first today's flash memory is NAND memory http://en.wikipedia.org/wiki/Flash_memory

NAND memory is written with tunnel injection, which causes charge carriers to be injected into a conductor. http://en.wikipedia.org/wiki/Tunnel_injection

Charge carriers "In semiconductor physics, the traveling vacancies in the valence-band electron population (holes) are treated as charge carriers"http://en.wikipedia.org/wiki/Charge_carrier

So we're using an electric charge, to fill, and create 'electron holes' in a conductor. what could POSSIBLY go wrong, in the real world, rapidly changing if a conductor has electron holes or not, by forcing the electrons in or out of the material ...

so trying to intentionally wear out a NAND memory chip can cause a severe problem whereby instead of creating an electron hole, you've created a short circuit. "A short circuit (sometimes abbreviated to short or s/c) allows a current to flow along a different path from the one intended." http://en.wikipedia.org/wiki/Short_circuit

Longevity

[Bananatree3]

I agree that high quality platter drives will last a long, long time. The issue is that anything with moving parts is inherently more prone to breakage than a device with no moving parts. A SSD with no rewrite issues would by principal be inherently longer lasting.

Platter drives are here to stay for a while. Once SSDs get the bugs worked out and the price drops to current platter drive levels, there will be a large migration.

Re:Longevity

[hairyfeet]

While I agree that the enterprise users will end up on SSD,simply due to IOPS, I'm not sure if it will end up in consumer rigs other than specialized laptop situations, and what is more I'm not sure I'd want them to be. We all know how the big PC makers end up starving the low end machines for RAM,so how well will a SSD survive say, a Vista Basic machine with 512Mb of RAM pounding the swap to kingdom come? I know they have wear leveling but consumers are a LOT more likely to fill their drives with junk and I bet that having to move the contents of the entire drive to even out the wear would kill any benefits you would see from SSD speed wise.

Also,what about data recovery? If a little Helen Homemaker gets a bug that bones her pc and she loses her pictures I can often get them back due to the fact that HDD drives just don't "erase" when something gets tossed. If the SSD does as I assume they do and actually erases on delete that could mean disaster for these customers that get virus infections, as for some reason lately they seem to hose the My Documents folder about 65% of the time and that is where many keep their pictures and other personal files,without a backup naturally. Can you even recover deleted files off a SSD? Or does the wear leveling software make that an impossibility? Anyway that is my 02c,YMMV

Re:Longevity

[hairyfeet]

What I meant was that thanks to the wear leveling going on under the file system there might be more of a chance of a file getting deleted from an empty sector as that sector gets overwritten by other files from elsewhere to even out the wear across sectors. Since many of these machines have or will have Windows on them,let me give an example using XP:

Let us say you have a 6Gb flash with XP installed. For ease let us say the drive is broken up into 12 sectors-A-L. The XP Install is on A-D and the file I delete is on sector E. Since I have not installed updates the OS has not really done much I/O since installation so the wear leveling moves my OS from A-D to E-H so that the sectors that were not used very often now come into play. Since the file that I want to recover was on sector E which now contains my System32 folder it will be a lot harder to try to get anything back than if it were like a HDD and simply left things where they lay. But I'll be the first to admit I haven't taken a close look at the algorithms used for wear leveling so maybe they have a way to recover those sectors,I just don't know which is why I asked. Anyway that is my 02c,YMMV

Re:Longevity

[orkysoft]

I think the harddrive would not know about file systems, and would actually swap the data between those sectors, but keep the old sector numbering, so it would be invisible to the higher layers, just like virtual memory works.

Re:But when

[pyite]

will they be competitive with mid range priced hard drives? You can get 500GB for $100 these days.

In a few years. Right now SSDs perform incredibly in terms of IOPS (I/O operations Per Second) that enterprise storage type folks are eying them longingly. They just need a little bit more space for the money. Until such time, it's very possible that we'll see more in terms of using SSDs as caching components in front of more antiquated spinning media.

Re:Yeah, Right

[Courageous]

Well. These drives (FC, SCSI, SAS) are 10% of the market, very lucrative, and quite important for data center operations, server rooms, and so forth.

Projected lifetime for modern SSD drives is now getting to the point where they are more likely to be discarded due to technological obsolescence than they are to significantly deteriorate, BTW.

The projected intersection curve is further than six years out for SATA SSD price parity. That's an eternity in technological time, which is to say, there is no predicting it.

Price per unit of storage is by far not the only deciding factor, even in the consumer market. Flash can scale up performance much more quickly than spinning media. You can expect flash performance to more than double annually from here on out, I would say. You would of course be right to be wondering how the SATA and SAS busses will keep up.

Look at FusionIO (http://www.fusionio.com) to see how flash will accelerate in performance. These devices have 160 internal channels in order to make the bytes flow at the rate they do. You can think of it as a sort of 160-wide RAID-0 striping mechanism.

$2400 for one card is of course way out of consumer space. However, point: 1) the cost of the flash in the system will drop to a fraction of its current price within two years, and 2) the ASICs on board this device will be "paid for" within the same period, allowing them to charge only a small fraction of their current price.

Expect other similar products to develop soon.

When FusionIO proves out the market for these devices--and mark my words, they will--competitors will follow in their footsteps, like bees drawn to honey.

C//

not only price, but density

[fermion]

If we reflect back on the floppy disks days, we see that it was not only cost, but density, that killed the floppy oh so many years ago. A floppy was no longer useful for installing apps. MS often needed upward of 10 disks to ship an app. While 3 MB was big enough to hold most files, we were entering a period in which one could no longer survive with a single 3.5" disk. The CD-R, then the DVD-RW, made sense as they could replace the floppy, though in many ways at a higher costs, due to their higher density. The fact that CD was cheaper than other optical solutions made it a good choice. What did finally kill the floppy was the available of USB drives for the sneaker net. Though expensive, they too had a density benefit, as well as not requiring additional hardware, other than a USB port which initially were scarce on MS Windows machines, and the drivers buggy.

I think that density, not price, is going to drive the SSD market as well. We need space on our small computers, and the mechanical solution is not keeping up. I believe this is why Apple went to flash memory for the iPods, although initially they were dedicated to hard drives. My iPhod mini only has 4 gb, the same as the nano that replaced it. The new nanos have more memory than even the EOL minis. The microdrive, though a good tech, were not scaling. The larger physical size hard disks are now up to 160GB, but that is small for modern times in which many of us have a terabyte sitting on our home machine.

So I think we will pay for SDD prices if they give us more space. The problem right now is that we have more for a SSD drive, and get less space. We pay $1000 to Apple or practically anyone else for 64GB SSD. That is paying money for nothing. Wait until we can buy a Macbook Pro with a terrabyte SSD for $4000, or a Mac Book air with a 250GB SSD for $2000. Then we will be seeing the SSD laptops flying off the shelf.

Of course for low end machines many will stick with HDD for many years, just like people entered the 21st century still storing things on floppy. Of course this will hasten the downfall of HDD, as the cheap unreliable HDD will take an even bigger share of the market than they have today, and, just like today, users will attribute a high failure rate to a problem with the technology, and not that they chose to buy a cheap hard drive. With the last major mechanical part gone, computer will become much more reliable, just like when the stereos, for better or worse, left vacuum tubes behind.

I also hope that DVD drive as a standard goes away soon, and applaud Apple for making the Mac book air drive free. The main reason for a dvd drive, other than installing software, is because we cannot rip out DVDs to a more convinent format. I would much rather carry around a couple Flash Drives than a bag of DVDs. It would seem that in not too many years shipping software on USB dongles would be just as cost effective. Already 4GB flash cost less than $10.

SSDs will reach price parity on June 15th

[DDumitru]

On June 15th, Mtron will start shipping the 1000 series MLC drives. Put these in an array with the right software and you end up with price/GB parity with 36GB 15K 2.5" SAS drives and about 12x the random IO performance.

HDD Array:

8 Seagate Savvio 2.5" HDDs: $350ea $2,800
    configured raid-10
1 SAS raid controller $600
Total cost for 144 GB $3,400 or $23.61/GB

SSD Array:

6 Mtron 1025-32 2.5" SSDs: $290ea $1,740
    configured raid-5
1 SATA raid controller $250
MFT Software License $1,250
Total Cost for 144 GB $3,240 or $22.50/GB

HDD Performance:
    4K and 8K read IOPS: 250/2000 (single-threaded/multi-threaded)
    4K and 8K write IOPS: 1200

SSD Performance:
    4K read IOPS: 8000/48000 (single-threaded/multi-threaded)
    8K read IOPS: 6000/36000 (single-threaded/multi-threaded)
    4K write IOPS: 40000
    8K write IOPS: 22000

These performance numbers are with the MFT driver in place. Without MFT, the 4K random write performance is about 140 IOPS (>250x slower).

Endurance for these SSDs in this configuration is good enough to overwrite the entire array with random data three times a day (500GB of random updates/day) for about five years.

These drives make a wicked mail server (EasyCo just moved one of it's mail servers mirrored to MLC flash and the difference is amazing).

Sorry for the blatant advert, but SSDs are here now.

Doug Dumitru
EasyCo LLC
http://managedflash.com/
+1 610 237-2000 x2