Leaked Intel Roadmap Shows 600GB SSD

An anonymous reader writes "Solid State Drives have been trying to fill the mechanical hard drive niche for some time now. The problem is that while flash memory is faster than a spinning platter, it is also much more expensive per gigabyte. Over the weekend details leaked about Intel's SSD roadmap, and what's most interesting about it is that the capacities of Intel's SSDs are going to increase in a big way. First off is a refresh to the high performance X25-M range of SSDs. Currently available in 80GB and 160GB models, these will be replaced by a new design, codenamed Postville, which will come in 160GB, 300GB and 600GB variants."

Comments

price still needs to come down!

[Joe The Dragon]

price still needs to come down!

Re:price still needs to come down!

[dgatwood]

I couldn't disagree more, for several reasons:

• Your OS, drivers, and applications will easily eat half of that 64 GB without saving a single file of user data.
• A web browser's on-disk cache typically hovers at another gig or so.
• My photo collection alone is 60 GB. Sure, I take lots of pictures, but as megapixel counts increase, the size of photo collections does, too. That's mostly from shooting at the smallest size on my DSLR....
• In this day and age, most computer users buy laptops as their primary machines because they are portable. External hard drives are the opposite of that.
• External hard drives drain your battery much faster than a larger internal drive. Much, MUCH faster.

My current laptop HD is 500 GB. I have only 50 GB free. Now about 240 GB of the space taken up is in the form of large files that could reasonably live on an external HD because I don't really need or want it with me. Still, that means I have 210 GB of stuff that I legitimately would want to carry around at all times, up from 160 GB two years ago when the last drive died, meaning that I pack on an estimated 25 GB per year of new material. And even that pales compared with people who do lots of movie downloading (legal or otherwise). (Yes, you could argue that those downloads could be put on an external drive, but that becomes a management headache when deciding what movies to bring with you on a trip, and... you get the idea.)

With the upswing in downloadable content (both movies and software), the need for hard drive space is in a rapid upswing. If most people only needed 60 GB drives, you'd still be able to buy spinning drives that small, and the few percent of users who needed the bigger capacity would have to deal by adding external drives. Since we're not seeing any sign of the demand for larger drives slowing, I think it's safe to say that 64 GB is not enough for most users. I doubt it is enough even for most casual users.

Re:price still needs to come down!

[DJRumpy]

I have to wonder why the cost is still so expensive. They are starting to see widespread use, with most vendors offing an SSD selection for notebook and desktop models. Most technology typically experiences a rapid drop in price long before the level of market acceptance we're seeing for SSD. These have been available for years now, yet the price is still prohibitive. Is it the raw materials that are so expensive? The R&D for the basic design is pretty much a done deal at this point, no?

Re:price still needs to come down!

[ddegirmenci]

Not to mention gaming...

Re:price still needs to come down!

[peacefinder]

Not really. (Different version.)

With a new SSD, one can sometimes remove a substantial performance bottleneck in an otherwise adequate older machine. Dropping a few hundred bucks on a new SSD drive might delay the purchase of a whole new machine by a year or two. From there, it's pretty easy to see why people wil be willing to pay pretty stiff prices for SSDs and also why Intel would be extremely motivated to not miss out on that market.

Re:price still needs to come down!

[fattmatt]

I've heard this before ... something along the lines of "64 GB ought to be enough for anybody"

Re:price still needs to come down!

[xavieldar]

Flash memory is actually very cheap when you compare it with other memories made of silicon. In term of density (bit/nm^2), there is nothing better today: a flash cell is approx 3 times smaller than a DRAM cell. Additionally, today, a single flash cell can store up to 3 bits (Multi Level Cell, MLC as opposed to Single Level Cell, SLC). However MLC-based flash memories require more time to read and to program pages and have a lower life time.

As the price of a chip roughly depends on its die area, it also makes flash the cheapest memory existing today. Today MLC Flash is ~11x denser than DRAM, while SLC is around 3x denser, that makes it approximately 11 times and 3 times cheaper than DRAM, respectively. Look at the prices of DRAM and Flash in your favorite store, you should get close to these ratios.

In the future prices should lower as we make flash even denser. However there is a limit to this shrinking: in 3-4 years the price reduction of flash should slow down. Another issue I heard about is that if everybody in the world wanted to replace their magnetic disk to SSDs today, there simply would not be enough silicon available...

Any update in terms of long run use?

[e065c8515d206cb0e190]

Not trying to be ironic here, but do we have any idea on how those will behave in the longer run? Are there improvements from the previous generations? TFA doesn't have much information besides capacity.

Re:Any update in terms of long run use?

[Kepesk]

Agreed, SSDs still have many cost and reliability issues to overcome, and I'm not going to get too excited till I see some improvements in those areas. Solid State is the wave of the future, but the wave is still way out there and is only just reaching the rocks off-shore.

Re:Any update in terms of long run use?

[FuckingNickName]

When I see real-world usage reports of SSDs under a range of regular HDD duty cycles, rather than hand-waving "well with the wear levelling algorithm you should get about xyz writes by which time you totally would have worn out your spinning rust" (oh, really?), I might consider applying them to servers which require frequent writes.

Re:Any update in terms of long run use?

[Anonymous Coward]

Frankly, mechanical hard drives are so hilariously unreliable as well that I don't see how it would make a difference: you need a redundant array with frequent offsite backups either way, right?

Re:Any update in terms of long run use?

[pezpunk]

yeah because mechanical hard drives never, ever crash...

Re:Any update in terms of long run use?

[ocularsinister]

Bear in mind that when a hard disk fails you typically loose at least some of your written data, and in worst case scenarios all of it. You won't be able to write to certain areas when an SSD fails, but you can often still read the data. So, yes, SSDs might fail a bit sooner, but its usually not critical like a hard disk fail.

Re:Any update in terms of long run use?

[Coren22]

Yeah...I am sure that you have looked at the reliability numbers...like ever...

Intel x-25m reliability: http://download.intel.com/design/flash/nand/mainstream/mainstream-sata-ssd-datasheet.pdf

BER (read error rate) of 1 sector per 10^15 bits read
MTBF 1,200,000 hours
Minimum 5 years useful life

WD Raptor Reliability: http://www.wdc.com/en/products/products.asp?driveid=495

MTBF 1,400,000 hours
Other figures not given

and the WD Raptor is considered an Enterprise hard drive, so that should say something about the reliability expected. I don't see these drives failing any time soon, and I have a Intel x-25m 32GB I bought a little over a year ago running quite strong with no errors in my desktop that rarely is shutdown.

The only reliability problems I have seen is in MLC based drives we use here at work for database servers, they go offline and have to be reseated in order to bring them back, but we haven't had any of these fail yet even under the heavy strain of a database server.

Re:Any update in terms of long run use?

[gumbi west]

That criticism makes sense for a netbook drive where when it dies you just replace it and no need to backup--the email are already on IMAP and everything else was just caches. But for places where you really care about your data then there are all sorts of other questions: how does it crash? Does it crash in such a way that the RAID you are using keeps its integrity?

In general, conservatives (in the sense of not wanting to change) are right to be conservative because of the long arm of the law of unintended consequences. People who try new things can end up with better results if things go as planed. But there are many more ways for things to go not as planed and for the project to crash and burn--leaving you at square one with nothing to show but lots of money/time spent on a cinder.

Re:Any update in terms of long run use?

[arth1]

Unfortunately, Intel seems to do like the rest, drop SLC in favour of MLC. That has a huge negative impact on both reliability and performance, but brings the price down and the capacity up.
That said, Intel's MLC drives are pretty good for MLC drives -- the X-25M is best in class, but still far below the speed and reliability of the X-25E.

If Intel could come out with a 128 GB X-25E, I would buy it immediately over a 600 GB X-25M at the same price. But they won't, because people don't want what's best, they want what's cheapest that still carries the "right" name.

Re:Any update in terms of long run use?

[rabtech]

Agreed, SSDs still have many cost and reliability issues to overcome, and I'm not going to get too excited till I see some improvements in those areas. Solid State is the wave of the future, but the wave is still way out there and is only just reaching the rocks off-shore.

That greatly depends on your specific application. I can tell you that installing an SSD in my work laptop was the single greatest (relative) performance jump I've ever seen, starting with my 8086/1MB/CGA machine until the present day, including all processor/memory/graphics upgrades I've ever done.

I can also say that some Antivirus products really, really suck and take up tons of CPU and have single-threading bottlenecks, so that if you have the RTV scanner turned on, you will give back a lot of the performance gains. (I'm talking about the one that installs 19 different drivers and services. Someone in IT got a kickback on that purchase).

I'd pit this SSD against a mechanical hard drive in a laptop any day of the week. It can take all sorts of bumps, bounces, heat, etc that could kill a HDD. Better battery life, increased performance. At 160GB, it is about 100GB less than the HDDs they are installing in new laptops, but other than that it is better in every way.

Re:Any update in terms of long run use?

[FuckingNickName]

The manufacturer data sheet is pretty much the polar opposite of "real-world usage reports... under a range of... duty cycles".

Re:Any update in terms of long run use?

[Rockoon]

I keep hearing people claim reliability issues when SSD articles come along to slashdot.

I have never seen a citation, so I went looking for them via Google but could only find citations attesting to the high reliability of these devices.

Dell's Lionel Menchaca stated in 2008, when it was reported by Avian Securities that Dell was having SSD reliability issues, "Our global reliability data shows that SSD drives [that we shipped] are equal to or better than traditional hard disk drives we've shipped." He further notes that Avian Securities never contacted them and that their numbers were a complete fabrication.

At this point I consider any claims that SSD's are less reliable to simply be a myth derived from dishonest reporting.

Furthermore, there are published studies detailing how unreliable traditional magnetic platter drives are.

Do they have write limits? Yes. Can other parts of the device fail? Yes. Are they more expensive than economy platters? Yes. Is there real world data showing that they are less reliable as claimed? Apparently not.

Re:Any update in terms of long run use?

[nschubach]

That's why I think hard disks will still be the norm for mid-term retention for a while. It can't take much to run your system off an SSD and mirror it to a platter... can it?

My best guess would be like a "hybrid" drive that uses the SSD for all immediate tasks and cache write that data to disk when it's free. In the event of an outage, you still have the data on the SSD which should always be considered accurate and you have the platters in case the SSD fails.

I'm pretty sure there are no RAID controllers that support that, but my RAID knowledge is limited to the basics of 0, 1, 5, 6, 10 so there could be...

Re:Any update in terms of long run use?

[mcgrew]

Bear in mind that when a hard disk fails you typically loose at least some of your written data

No you don't, your data is NOT loosed, it's locked up so tight even you can get at it. You loose your data when you publish it, you lose your data when your hard drive dies.

Hope I was a help. What's your native language?

Re:Any update in terms of long run use?

[networkBoy]

FWIW I've already seen flash device failures (SSD and USB sticks). They tend to fail into a RO mode rather than a blank, or unreadable mode. This is a good thing from a data integrity standpoint (though a bad thing from an IS standpoint).

I personally would feel comfortable using SSDs in a transaction server and such from a data integrity view, but I'm not sure if they could actually handle massive IOPS for a sustained period. Massive OPS, however, they seem to be awesome at, and that's how we're currently using them. A front end cache for largely static datasets that need high read availability. Where we used to be bottlenecking on the drive, we are now bottlenecking on the controller logic.

Re:Any update in terms of long run use?

[bobcat7677]

Certain models of Adaptec controllers with recent firmware (since April 2010) support SSDs and platter drives on the same mirror array (RAID 1 or 10). The controller intelligently sends all reads to the SSD unless it goes offline. It's not at all an advertised feature, I have only ever seen mention of it in the firmware release notes. Note that this is not the same thing as what their "MAXIQ" product does, which is essentially add more cache to the controller in the form of a small SSD attached to one of the controller's ports.

Re:Any update in terms of long run use?

[Kjella]

Unfortunately, Intel seems to do like the rest, drop SLC in favour of MLC. That has a huge negative impact on both reliability and performance, but brings the price down and the capacity up.

I've never heard that reliability is that much different, but durability yes. SLC drives can take about 10x as many writes per cell before they wear out. However, MLC drives are rated at 10000 writes/cell too and smart algorithms avoid overusing single cells. Each MLC cell is slower individually, but by writing to many in parallel both have IOPS way, way beyond traditional drives and we're discussing degrees of lightning fast. In other words, both the shortcomings are largely avoided by making smarter controller chips. Unfortunately due to the latter the small and cheap SSD never really came because few cells mean low speed. So when Intel announces bigger drives, it probably means faster too. But compared to a HDD, SSDs are already plenty fast. Try trashing your drive with some random writes and a HDD will grind to a halt, even my Vertex beat the fastest HDDs by a factor of 10x under those conditions.

Re:Any update in terms of long run use?

[fnj]

My guess is that O.P.'s native language is English. I looked at a bunch of his posts, and he seems very fluent and adept. I think you will find O.P. is an example of a very intelligent and capable individual who is the product of a badly failed educational system (in this particular example, failed teaching of English language is noted). Hardly any memorization is taught any more. The rule that "oo" is pronounced as in "ooze" and "o" is pronounced as in "foe" is successfully taught, but the table of exceptions to the rule is not taught successfully at all. Judging by the results, it is not taught at all, or only very peremptorily. Geography and history are two other subjects the scope and quality of whose teaching has dropped to a very low ebb. I think you will find a far better quality of English teaching in India than in the U.S.

Admittedly we are both fixating on a single misspelling, but in the case of this particular single misspelling, what was for a great many years successfully taught, has in fairly recent years become almost universally not so. The word is misspelled almost as often as it is correctly spelled.

Check out the 8th grade final exam from 1895 from Salina, Kansas. This is just the grammar portion:

1. Give nine rules for the use of capital letters.
2. Name the parts of speech and define those that have no modifications.
3. Define verse, stanza, and paragraph.
4. What are the principal parts of a verb? Give principle parts of "lie," "play," and "run."
5. Define case; illustrate each case.
6. What is punctuation? Give rules for principle marks of punctuation.
7. Write a composition of 150 words and show therein that you understand the principle rules of grammar.

And a portion which was called "orthography:"

1. What is meant by the following: alphabet, phonetic, orthography, etymology, syllabication?
2. What are elementary sounds? How classified?
3. What are the following, and give examples of each: trigraph, subvocals, diphthong, cognate letters, and linguals?
4. Give four substitutes for caret "u."
5. Give two rules for spelling words with final "e." Name two exceptions under each rule.
6. Give two uses of silent letters in spelling. Illustrate each.
7. Define the following prefixes and use in connection with a word: bi, dis, mis, pre, semi, post, non, inter, and mono.
8. Mark diacritically and divide into syllables the following, and name the sign that indicates the sound: card, ball, mercy, sir, odd, cell, rise, blood, fare, last.
9. Use the following correctly in sentences: cite, site, sight, fane, fain, feign, vane, vain, vein, raze, raise, rays.
10. Write 10 words frequently mispronounced and indicate pronunciation by use of diacritical marks and by syllabication.

Even though the above does not much demonstrate very specific memorization, I believe it serves to make the point that standards have fallen very far. Not only could virtually no present day U.S. 8th grader pass that test, I submit few college seniors could; even those who major in journalism.

The geography section is also an eye opener:

1. What is climate? Upon what does climate depend?
2. How do you account for the extremes of climate in Kansas?
3. Of what use are rivers? Of what use is the ocean?
4. Describe the mountains of North America.
5. Name and describe the following: Monrovia, Odessa, Denver, Manitoba, Hecla, Yukon, St. Helena, Juan Fernandez, Aspinwall and Orinoco.
6. Name and locate the principle trade centers of the U.S. Name all the republics of Europe and give the capital of each.
7. Why is the Atlantic East coast colder than the Pacific in the same latitude?
8. Describe the process by which the water of the ocean returns to the sources of rivers.
9. Describe the movements of the earth. Give the inbclination of the earth.

Beh

[Pojut]

The price is still far too high. I recognize that an SSD can provide a good performance boost, but still...the prices are way too high. I'll likely give it another year or two before I pull the trigger on one.

Not that any of you care -_-;;

Re:Beh

[Pojut]

Agreed...if they keep the price points the same, but double capacity, I would be much more inclined to pick one up. I know you don't technically *need* alot of space for a system drive, but I don't like having such limited free space. 160GB would be the absolute bare minimum I would use for a system drive these days, and even that's kinda pushing it.

Re:Beh

[EvanED]

BTW I don't want to say "oh, everyone will have huge games like this" and whatnot, but at the same time, it's also way of an overgeneralization to say "80 GB is plenty".

And to continue my last thought, you could say "just put what you need" on the SSD, but that presents its own problems. How do I decide what to put on there? Do I need to be installing and uninstalling programs as I change which ones I use more? How much more of a pain is this with Steam, where you can't choose an install directory? (BTW, are you listening Valve? Add this feature.)

These sound like a huge pain, which is why I'm holding off on an SSD for a little while longer. When I can have a magnetic "media" drive for huge stuff that doesn't need fast transfers (videos, rips of my CDs as FLAC, etc.) but have an SSD for *all* or basically all my programs and most small personal data, I'll get one. In the meantime, even though I do want one, they basically seem like they'd be a bit of a pain. (I'd guess the former will happen in more than one and less than two years, but we'll see.)

"Postville" is the current generation

[Florian Weimer]

For example, this is a posting using the code name: http://communities.intel.com/message/51359;jsessionid=F3036FCC8C1DD878FCED25A7A6D32547.node6COM

Re:"Postville" is the current generation

[malzfreund]

That's correct. It would be more accurate to call it the Postville refresh (which uses 2Xnm NAND Si). "Postville refresh" is the term Intel uses on one of the slides that leaked.

Re:"Postville" is the current generation

[malzfreund]

No, the OP clearly refers to the Postville refresh, which will bring capacities of 160/300/600GB NAND. Lyndonville is the codename of the follow-up to Ephraim, i.e., Intel's series of enterprise drives commonly known as X25-E. Lyndonville is expected in capacities of 100/200/400GB so that's clearly not what the article referred to.

No mention about speeds

[rsborg]

Intel does not have the fastest MLC drives out there (X25-E is SLC), and now they're ditching SLC?
I wonder how their performance will match the other controllers (Sandforce, Indilix, Samsung, etc)... perhaps their new MLC is more along the lines of what Sandforce is doing?

Re:How will large SSDs effect databases?

[Surt]

People who tune large databases have been IOPS focused for a long time. SSDs enable a new level of IOPS that is about one to two orders of magnitude better than spinning disks. SSDs will allow people to (re)consider all sorts of applications that are currently IOPS bound or IOPS prohibited. Soon Google will be able to keep track of how much milk you have in your fridge, and send you a reminder to buy some when you are near a store that sells it, and have plans to go home afterward so that they can be sure you will be able to refrigerate it.

http://online.wsj.com/article/SB10001424052748704901104575423294099527212.html?mod=WSJ_Opinion_LEADTop

What the?!

[Sits]

While I can somewhat agree with your sentiment (64GBytes isn't a lot when you are saving media data) I feel you have exaggerated a bit in the OS numbers:

• The OS I'm typing this on (which is on a Intel Core 2 laptop with 4GBytes of RAM) is taking up 6GBytes and has various development tools and libraries installed on it. The OS on my EeePC takes up 3GBytes.
• Even on the bigger computer the current Chromium cache size is 437MBytes. Perhaps it scales with disk size?

On all but the most unusual of setups (I know people who do FPGA development whose tools take up 20GBytes by themselves) it's going to be "user data" that is taking up the vast majority of the disk space - not the operating system and applications (given that most operating systems still ship on no more than a single 4GByte DVD you would need compression of about 8:1 to fill up the disk from that alone). I have no doubt that if you take photos or have a big movie collection 500GBytes is not going to see like all that much though.

Re: vs. Spinning Platter

[PitaBred]

Very much so. But hard drives with the shock protection are still pretty robust. I love having the SSD in my machine... it's amazing how fast everything goes. Programs start instantly, it boots so fast that I disabled hibernation, but I'm still at a paucity of space with a 256GB SSD.

The thing you're paying for with SSDs is performance. If you haven't used one, you don't know what you're missing, but if you have, you never wanna go back to things the way they were.

The price probably is coming down, probably half

[TravisO]

Knowing Intel, most likely the new drives, which double the capacity, will remain the same price. Therefore your 160gb drive for $399 will be 320GB for $399, and imho approaching $1 per GB on SSD is a big freaking deal. Of course this info is an early leak and Intel has no mention of price. But with a new fab and smaller nm, most likely Intel will deliver on this theory.

Re: vs. Spinning Platter

[vadim_t]

Not as robust as you might think.

Shock protection (unless there's some development I'm not aware of) measures acceleration and parks the drive's heads if the acceleration is too much, in case that acceleration means the laptop is about to hit the floor. It's a good idea, but the application is limited. It's excellent when you drop the laptop, but it won't do you any good if you give the laptop a good jolt without any warning.

I killed my laptop's drive once by turning around in an office chair, and hitting the laptop with the chair's back. Drive immediately started making weird noises, and I spent all night copying stuff off it. The acceleration sensors are completely useless for something like that, since there's no way for it to guess an impact might be coming.

Re:Troll feeding time!

[EvanED]

A quick scan of Newegg shows that a SDD costs ~$2.21/GB, where a comparable traditional HDD costs only ~0.33, thats quite a difference, I'm not sure if 15 minutes of battery life, and perhaps (very generously) a second a day in seek/read/write time is worth that much.

I'm generally somewhat with you... waiting for SSDs to halve in price another time or two before I jump in.

That said, you're underselling SSDs a lot here -- a second per-day is not even remotely realistic in terms of saved seek time, and that even ignores the fact that good SSDs now beat most hard drives in raw transfer as well.

There are multiple people out there (including Linus Torvolds, Jeff Atwood, and some random poster in this story) who say that changing from a magnetic hard drive to an SSD is about the biggest single upgrade you could make to a reasonable system today. The random /. poster I mentioned said that upgrading to an SSD was the single biggest speed increase of any upgrade he's ever done. Of course YMMV and this is workload-dependent, but don't understate the benefit of a good SSD either.

To cover all of this with an SSD would cost significantly more than my full computer (which isn't a slouch hardware wise).

Of course, you wouldn't do this; you'd come up with some split between what should be on fast SSD and what should be on a slow magnetic media, and have one drive for each.

Re:Troll feeding time!

[EvanED]

There are multiple people out there (including Linus Torvolds, Jeff Atwood, and some random poster in this story)

BTW, if you want citations:

Linus on his Intel:

In fact, I can't recall the last time that a new tech toy I got made such a dramatic difference in performance and just plain usability of a machine of mine. ...
Everything performs well. You can put that disk in a machine, and suddenly you almost don't even need to care whether things were in your page cache or not. Firefox starts up pretty much as snappily in the cold-cache case as it does hot-cache. You can do package installation and big untars, and you don't even notice it, because your desktop doesn't get laggy or anything.

Jeff Atwood (admittedly, where I saw Linus quoted):

And, frankly, I was blown away by the performance difference compared to the 300 GB Velociraptor I had in my system before. That drive is not exactly chopped liver; it's incredibly fast by magnetic platter drive standards. ...
In my humble opinion, $200 - $300 for a SSD is easily the most cost effective performance increase you can buy for a computer of anything remotely resembling recent vintage. Whether you prefer the 80 GB X25-M SSD or the 128 GB Crucial SSD, it's money well invested for people like us who are obsessive about how their computer performs.

Trust me, you will feel the performance difference of a modern SSD in day to day computing. That's far more than I can say for most of today's CPU and memory upgrades. The transition from magnetic storage to solid state storage is nothing less than a breakthrough.

Random /.'er rabtech:

I can tell you that installing an SSD in my work laptop was the single greatest (relative) performance jump I've ever seen, starting with my 8086/1MB/CGA machine until the present day, including all processor/memory/graphics upgrades I've ever done.