Intel-Micron Joint Venture Develops 25nm NAND 121
Ninjakicks writes "IM Flash Technologies is a joint venture between Intel and Micron that is targeted for producing NAND flash memory. With a focus on R&D, IMFT has doubled NAND density approximately every 18 months. Tomorrow IMFT will announce the launch of their 25 nanometer NAND technology — a major advancement in the semiconductor industry. Intel and Micron can now lay claim to the smallest production ready-semiconductor process technology in the world. IMFT took members of the press on a tour of the new 25nm fab and it's an interesting view into this bleeding-edge manufacturing process."
Great News (Score:4, Interesting)
This is sorely needed to bring down costs for SSD's. The price and capacities available are coming down at a disappointingly slow pace.
Re:Great News (Score:5, Informative)
I think more troubling in the SSD market has been poor design at the low end (see http://www.anandtech.com/printarticle.aspx?i=3531 [anandtech.com] for more detail). 150 dollars for a 64 GB SSD is fine, but when random write speed is an order of magnitude slower than a standard hard drive that costs an order of magnitude less, something is severely wrong.
Early adopters such as myself got pretty screwed over. Until consumers can trust the technology, I don't think price matters. Manufacturers need to put effort into building a high quality product first - they need to design good controllers and firmware.
Re:Great News (Score:5, Interesting)
Those JMicron drives were absolutely horrible - 4KiB of SRAM cache, later doubled to 8KiB? Even Intel's lowest end SSDs have 256KiB, plus another 32MiB of RAM for caching the locations of free spots to write files.
Oh, btw - the cache has to be SRAM so that if the power goes out, it can write the files when it comes back on. SSDs absolutely must have a RAM cache so that they can efficiently locate places to write files, or they will stall while the controller tries to locate one. That's why the low end controllers perform so horribly in random write.
But now even the worst controllers aren't too bad. If I remember right JMicron's newest low-end controller has 128KiB of cache, and there are cheap Intel knockoff SSDs coming out that perform decently. (same controller, but less RAM cache and less space) If what you want is blazing fast loadtimes in games, they aren't bad options, but they're still slower than a fast HDD and have way less space.
Re: (Score:2)
http://ncix.com/products/index.php?sku=49278&vpn=OCZSSD2-2SLD30G&manufacture=OCZ%20Technology&promoid=1210 [ncix.com]
30GB OCZ Indilinx drive for a little over $100 CAD. Not a bad option, but not much space.
Re: (Score:2)
As I found out when running Windows 7 on a Patriot Warp v2, 30GB is not enough. Unless you spend some time stripping
the OS down to barebones, you'll quickly run out of space. And forget about hibernate if you have lots of RAM.
Even with fast SSD booting, there are times when you'll want to save your programs state rather then start it up all over again.
Re:Great News (Score:5, Funny)
Fighting for peace is like fucking for chastity
This has always bothered me. Do you know of another way to produce more virgins?
Re: (Score:2)
Don't run your OS off it then. Install three or four games that you play quite a bit.
It is an option. You just have to factor the lack of space in.
Re:Great News (Score:4, Informative)
It would help if Windows would allow you to put the hiberfil.sys file on a different drive but you can't even move it to a different partition on the same drive.
Re: (Score:1)
http://www.mydellmini.com/forum/windows-7/2441-windows-7-ultimate-solid-state-drive-speed-tweaks.html [mydellmini.com] has a list of tweaks for windows on a SSD. Amongst them is disabling hibernate on the assumption that you're start up time is good enough.
Disclaimer: I've came across this researching SSD's before I buy & haven't tried it yet. Thought it might help you.
Re: (Score:2)
Thanks but I've done all of these before as these were all things tried to alleviate the stuttering problem with the notorious JMicron controllers - among other things.
My point was that, regardless of SSD speed, there are times when hibernation is very useful.
In the course of a browsing weekend, I may have up to 100 tabs/windows open as well as other programs.
True, I can easily save the session and/or restart the programs but with trying to reload so many tabs and restoring the state of a half-dozen or more
Re: (Score:1)
Obviously you have never installed Crysis...
Re: (Score:2)
No, I only play fun games. :P
I figured, one 12GB game, two 6-8GB games, and a 2-4GB one. But I suppose depending on what you play, it could be 2 to 150 games.
Re: (Score:2)
Re: (Score:2)
1. Hmmm, will look into that - I'm using Ultimate on that machine.
2. This is a notebook, with spare battery. Only space for 1 disk.
3. Once I'm sure an update didn't break something, I delete the backup folders.
4. I'd done that when I tried the Win7/SSD combo on a desktop,which was ok. But, I really wanted this for the laptops.
5. Yah, I learned years ago about System Restore invisibly consuming space. I usually turn it off or reduce it to 1 or 2%.
Good suggestions overall but my real wish was/is to have my la
Re: (Score:1, Offtopic)
Oh, btw - the cache has to be SRAM so that if the power goes out, it can write the files when it comes back on. SSDs absolutely must have a RAM cache so that they can efficiently locate places to write files, or they will stall while the controller tries to locate one. That's why the low end controllers perform so horribly in random write.
lol wut.
Re: (Score:3, Informative)
SRAM [wikipedia.org]. I do not think it means what you think it means.
Re: (Score:2)
Poor wording on my part.
Before I got a UPS, I had my power cut out or surge from time to time. On two separate occasions files were being written when my computer shut off, and got stuck in a strange not-a-file not-a-dir state. The files and their parent folders could not be deleted, but I could format the partition to clear them. One happened to be in my temp folder, which is how I first discovered the issue.
As long as power is restored shortly, SSDs are supposed to complete the write... or is that only se
Re: (Score:2)
Server-grade disks are intended for use in RAID configurations, so they need the cache to be battery-backed because otherwise they suffer from the write hole problem (write to one disk, power goes out, stripe is inconsistent). Often, the server grade disks will have no cache at all (which makes them a terrible choice for home users), because the server grade RAID controller will have a big, battery-backed, cache.
All drives, mechanical and SSD, exhibit the problem you are describing. This is why we've h
Re: (Score:2)
All drives, mechanical and SSD, exhibit the problem you are describing. This is why we've had journaling filesystems for the past decade or so.
Doesn't seem to have helped my NTFS partitions. ;)
Re: (Score:2)
Well that explains it. Cheers.
Re: (Score:2)
SRAM is not persistent, it doesn't survive power outages.
It's called 'static' because it doesn't need to be refreshed.
Re: SRAM caches (Score:2)
Oh, btw - the cache has to be SRAM so that if the power goes out, it can write the files when it comes back on.
That requires a battery or supercapacitor on the drive itself, and while some (expensive) SSDs have the latter, the feature is still relatively rare.
Of course any serious ("enterprise") application should have a UPS, but uninterruptible power supplies (or the power distribution networks) can and do fail from time to time, for a variety of reasons, so adding a supercapacitor to protect the (block le
Re:Great News (Score:4, Insightful)
I think in that case, the hardware vendors simply released a product that wasn't ready for prime time. Possibly to try to recoup R&D costs. I can't imagine that they weren't aware of the performance issue with invalid page data.
Larger capacities made possible by this 25nm technology should help to alleviate the issue somewhat, or at least extend the time before it becomes noticeable. That plus support for the TRIM command should go a long way towards deflecting the issue.
Re: (Score:2)
Near constant studders and stalling? Yeah, I'm sure they weren't aware. My dad bought me an OCZ Apex and it was worse than useless with XP. I have an SSD with a Samsung controller, now. It rocks.
All those USB key drives that fail after a few weeks of usage? ATI drivers with crappy installers? DRM that refuses to work after you use msconifig? Obviously the technology just needs to mature before they get the kinks out.
Re: (Score:2)
That part is mostly fixed, I got a OCZ Vertex (Indilinx controller) and I don't notice any problems. Neither do those with Intel SSDs, as far as I know. The biggest issue is that I checked, and right now it costs more for the Vertex than I paid in April last year. If they could say halve the price, then 30-40GB SSDs would be in the "normal" price range for laptop HDDs, only being much smaller. Then you could talk to people about big vs fast, because it wouldn't add such a huge premium to the cost which is a
Re: (Score:2)
Things have really changed since that article was published 10 months ago.
Im pretty happy with my OCZ Vertex 60gig on win7. It doesnt have the problems the jmircon based drives have and the prices on them are pretty good. I just saw the 60gig model go for $179 after rebate. Write/read speeds on Win7 are excellent even after you fill up the drive because win7 and the 1.4 firmware support TRIM natively. Now there's a vertex turbo model that gives a 10-20% performance increase.
Its funny how the market is pla
Re: (Score:2)
Well. I'm sorry if this is a not-nice comment, but it was pretty obvious as the market developed that the early versions just weren't ready, and what the probable pit falls of the technology were likely to be. I.e., I don't think it took hind sight to know that early adoption was risky.
Re: (Score:2)
Hopefully it's getting better. I bought a couple of these [kingston.com], 128GB model, and so far I've not been disappointed with read or write speed (in fact it's been snappier than the Seagate I had before). Granted, I'm not running a busy server. As for it not supporting TRIM, oh well, it's half-to-a-third the price of a Corsair, and WinXP which is the OS I put on one of them doesn't even support TRIM.
Re: (Score:2)
150 dollars for a 64 GB SSD is fine
For me the problem isn't the $150 but the 64GB. I'd probably be willing to spend up to $200 for a reasonable SSD for my netbook (which has a hard drive) but I probably won't buy one until it's closer to a dollar a gig, so I can get the space I need to actually do stuff. Performance is nice, but if it's 2-3x faster (real-world), it's worth 2-3x more, not 10x more (current laptop drive to SSD price ratio @160GB @Newegg). That's speaking as someone for whom speed is nice a
Re: (Score:2)
Nice sentiment, but wrong. A car example: high-end sports cars are about 2-3x faster than a Tata, and cost well over 10x more.
From a business perspective, it can sometimes be worth spending a large amount of money in order to gain a very slight performance advantage.
Re: (Score:2)
What's happened is that demand has been higher than expected.
Re: (Score:2)
Given that Intels SSD's don't seem to exist at the moment (the X25-E has been unavailable from major European distributors since 10/2009) I'd just like them to make some.
Only 25nm? (Score:5, Funny)
If it's only 25nm, how do you use your hand?
Oh! NAND. Gotcha.
Cheap SSDs in my lifetime? (Score:2)
I want a cool, quiet 300G for 200 dollars. Imagine....a computer needing to cool only the CPU/Chipset.......I can only dream.
Re: (Score:3, Interesting)
SSD's aren't going to be cheap soon, they have enough advantages over rust that they'll be an overpriced alternative until we stop using rust completely, which is still some time off.
You'll probably never buy a new 300G SSD for 200. You might buy one of a much larger size for $200 because by the time it happens we'll be using MUCH larger drives.
Re:Cheap SSDs in my lifetime? (Score:4, Informative)
Integration and hardware costs keep the cheapest hard drives somewhere around $40 (for new stuff). Someone will be tempted to play in that space with a $20 SSD, at which point people will get out their fingers and determine the per GB cost of the $20 drive and be very unhappy with larger drives that cost much more than that.
Also, in 2007, they were ~$7.50 per GB:
http://www.engadget.com/2007/04/25/ssd-prices-in-freefall-wont-overtake-hard-disks-anytime-soon/ [engadget.com]
Vs less than $3 today (just google it). So the prices haven't come down quite as much as the article predicts, but there are 11 months left in the year, and I made that calculation using an intel drive (which probably carries a slight premium to the market).
Re: (Score:3, Interesting)
Re: (Score:3, Informative)
"No one will ever need more than 640kb."
Your assumptions are just as correct.
SSD has one significant advantage: Moore's law, so every 18months (roughly) the size doubles or price halves.
That means, there will be soon 300Gb SSD for 200$.
There won't be a premium on SSDs, they are REALLY expensive to produce, that's why the cost so much, not due to a premium. You could call recouping R&D as a premium how much you like, but it's not a premium, they do not need to cover that cost as well, otherwise no R&
Re: (Score:2)
Kryder's Law applies to hard drives but the real advantage of SSDs is access time and throughput relative to capacity. Hard drives suffer from a worsening density to throughput ratio because the later only scales to the square root of areal density. SSDs have not reached that point yet but they will. As chip capacity goes up, it will take fewer chi
Re: (Score:2)
I want a cool, quiet 300G for 200 dollars.
For anyone unfamiliar with SSD drives, they are indeed completely silent, but they're definitely not cool. Perhaps best described as moderately warm to the touch. For many, that could translate as "your notebook will still feel too hot".
Re: (Score:2)
What SSDs are you running? I've a Patriot Warp v2 32GB, an OCZ Solid 60GB and a Kingston V-series 128GB - all run very cool.
If anyone using one of those in a notebook feel that is too hot, they must be from the planet Hoth.
Something other than NAND? (Score:2)
Could this process be used to build, say, CPUs?
Re: (Score:2)
Re:Something other than NAND? (Score:4, Informative)
Fair warning, I didn't read through the whole article, but in general:
No, this won't be directly applicable to CPUs. Microprocessors and memory are two vastly different beasts, on the manufacturing side. Memories are arrays of of the same thing, over and over - neatly organized, same size devices, requiring the same power supply and same operating characteristics. Microprocessors have many different structures, different size transistors for different things, different power supplies, different signaling levels to turn on some transistors and not others. The relative simplicity - really, the relative uniformity - makes memories easier, because you don't have to worry (as much) about balancing the effects of the shrink and the method to shrink across several elements. What's good for some might be bad for others, so the fewer elements you have, the more leeway you have.
That's not to say that this won't have anything to do with CPU development, it's all very inter-related But you can't just start making microprocessors of this dimension because you have working memory.
Re:Something other than NAND? (Score:4, Informative)
Much more to the point, NAND can tolerate a very high defect rate in the individual cells, whereas a CPU can tolerate almost none (with some defects you can disable a core or some of the cache and still salvage the part). Further, NAND gates operate much much slower than CPU transistors and their operational results are checked against error correcting data. A CPU transistor doesn't have that luxury.
Re: (Score:1)
Re: (Score:1, Informative)
Re: (Score:2)
Re: (Score:2)
32GB MicroSDHC (Score:2)
Re: (Score:2)
Re: (Score:3, Informative)
Probably soon enough. But if I recall correctly, 2GB was the maximum for SD, and 32GB is the maximum for SDHC cards. After that you need SDXC.
Re: (Score:2)
A lot of SD only devices choke on 4GB SD cards. Was the interface specification vague?
It seems incredible to me that the original SD card specification did not look far enough forward. Twice. Does every generation have to learn this lesson?
Re: (Score:2)
The way I'd do it is as follow: ask the SD card what capacity it has. The SD card returns a string of 0's followed by a 1 to end the string. The number of zeroes equals to the number of bits for addressing. Same goes for sector size (one zero = 256 bytes), etc. That way there's no lower or upper limit, and all SD devices (if programmed correctly) would be able to read all SD cards, even a reader from 2005 would be able to read an SD card from 2050.
Please, (Score:1)
Double patterning has limits. (Score:5, Interesting)
in the olden days, xx nm really meant feature size. With Intel and other fabs pressing mfg to half the size every 2 years, it seems mfg has gotten quite creative in their definition of feature size. Latest feature size is a fraction of the wavelength of the light used for patterning, and to achieve it, double and sometimes triple patterning is used. That is basically multiple exposures with slight offsets. The result migh be called 25nm but might really be 50nm, and edge sharpness when you are at 1/4lambda is so suspect that you really have to add some margins here and there, and some features dont really lend themselves to double and triple patterning, so you really have a mix including 50nm process for these.
Kind of like a marketing gimmic, just here it is engineering selling it as 25nm to their own marketing departmens.
Re:Double patterning has limits. (Score:5, Informative)
Re: (Score:1)
Re: (Score:1, Interesting)
The practice of optically shrinking an existing die without redesign/relayout is known as half-node stepping. If you read the analysis of these 25 nm parts over at Anandtech [anandtech.com] you will see that this is clearly not a half-node step. These parts are running charge trap memory cells whereas the previous generation used floating gate cells. Personally, I'll take the increased storage density any way I can get it.
Re: (Score:1)
Re: (Score:1, Informative)
I think he's saying that if your part has 10% 25nm features, 30% 32mn, and 60% 45nm; that's not really a true "25nm" part :)
(ie because some of the structures you need to etch have concave or oblique geometry so that a multi-exposure of part of the structure would ruin adjacent parts (capacitors and stuff, I guess)).
Micron? Seriously? (Score:2)
I hadn't heard anything about them in perhaps a decade. I can't believe they're still around! Didn't they used to have a desktop line too? Or am I thinking of someone else altogether?
Re: (Score:3, Informative)
Wikipedia to the rescue:
http://en.wikipedia.org/wiki/Micron_Technology [wikipedia.org]
They split their PC manufacturing business into a spearate company, which declared bankruptcy in 2008. Now, they focus on manufacturing memory.
To most of the people, Micron is known as their consumer brand Crucial Tehnology and Lexar Media.
Re: (Score:2)
They had a PC subsidiary, yeah: Micron Computers, from 1995 to 2001. They spun it off in 2001, and it continued under the name MPC until it went out of business in 2008.
It was never a huge part of their business, though. Micron's a large semiconductor company, and been a dominant player in memory chips for decades. The other stuff they've dabbled in --- consumer PCs, motherboards, briefly video cards, etc. --- seems never to have taken off.
8GB per chip (Score:3, Interesting)
---
Storage [feeddistiller.com] Feed @ Feed Distiller [feeddistiller.com]
Re: (Score:2)
That not bad storage per chip. Now they need to be able to pack 16 of them into a standard flash stick, for 128GB flash sticks. I'll bet they top out at 64GB per stick though.
If you're willing to pay, there's already a 256GB memory stick [kingston.com], I see it in stock for about 800$ + VAT here, so I guess this makes it possible to go to 512 GB. Not that I really see the sense in this product given the access speed, but I guess it's good for bragging rights.
Re: (Score:3, Interesting)
Withstanding recently trends in the flash market pointing to a slow down of the very fast price drops that have been happening, flash will beat 15K drives on price within 2 years or so. SATA is 6+ years. That may as well be an eternity in technology time. All bets are off. By then, one of the platter manufacturers could pull a density rabbit out of the hat.
Re: (Score:2)
All bets are off. By then, one of the platter manufacturers could pull a density rabbit out of the hat.
But would that even matter? For daily use 200GB have been more then plenty enough for me and that for quite some years. Sure I might have an additional 1TB drive for a video collection, but the core drive in the PC where OS data is stored doesn't really need to be super huge, 256GB at a good price should be enough for almost everyone and if the software would improve to allow caching/swap of frequently used HDD data on the SDD it could even be much less.
At this point I think raw price of the device is more
Re: (Score:2)
You're question "would that really matter" is a good one. You are right, in the consumer market things are judged on "price per," but rather "just price". In the enterprise mass storage market, however, yes, it will matter quite a lot.
This is not to say that there might not be an inflection point sooner than raw $$/GB. Savvy operators may have TCO models that switch them over sooner.
Joe.
Re: (Score:2)
I have not seen any slow down in hard drive capacity increase since GMR heads were introduced. Track density might have been a problem going forward but dual stage actuators are being introduced.
1 TB USB Flash drive! (Score:1)
I really would like a 1TB USB flash drive so i can carry around my pr0n collection everywhere and use my new iPad to view it all when I am on vacation!
Micron? (Score:2)
Re:Rall's Law intervenes (Score:5, Funny)
Re: (Score:1, Funny)
Actually, it is the case. You just have to reverse the polarity of the neutron flow.
Re: (Score:3, Funny)
Re: (Score:2)
Generating polarized neutron beams in a standard hard disk form factor is probably not practical.
Re: (Score:3, Insightful)
Re: (Score:2)
No good, if you shrink the neutrons they turn into neutrinos and leak out all over the place.
Re: (Score:2)
Yes, but you have to be careful and *not* cross the streams.
Re: (Score:3, Funny)
Especially Realplayer streams.
Re: (Score:2)
Thats not the case if you use a high-fraternity flux capacitor. I've heard the prazatonic implementation of intergalactical space bizanium will allow you to propate the quadraceptic meld process down to 0.2nm.
Just like over-inflating a balloon....
Re:Rall's Law intervenes (Score:5, Funny)
I hate to correct you here, but equilibrium is dependent upon factors of e, therefore the propation must be the 4th Fibonacci value derivation of the meld process.
You're joking, right? I thought everybody knew that Fibonacci values were bilaterally stochiocentric and corrupted the Van der Waals stress index at the third integral. How are you going to poststratify the gnomon clatch that way? Gods you're such a n00b.
Re:Rall's Law intervenes (Score:5, Funny)
How are you going to poststratify the gnomon clatch that way?
I use a hammer for that.
Re: (Score:2)
Whitworth Hammer?
And by the way... (Score:2)
I use a hammer for that.
The Hammer is my Penis. [tvtropes.org]
Re: (Score:2)
42.
Re: (Score:1)
http://www.youtube.com/watch?v=y2v6rXs5J9M
Bounce a graviton particle beam
off the main deflector dish
that's the way we do things lad
we're making shit up as we wish.
The Klingons and the Romulans
pose no threat to us
'cause if we find we're in bind
we just make some shit up.
Re: (Score:1, Funny)
Don't worry, as we speak Rambus is quietly patenting the shit out of everything in this space. A few years from now their patents will pop-up and bring Intel to its knees.
Re: (Score:1)
Re: (Score:2)
I strongly doubt that... x86 has so much cruft associated with it, it will never be able to hedge ARM out of the market, esp since the later is so entrenched at this point. Do you think cell phone designers want to work with the PCI bus? Chipsets? And what software is available for a non PC-compatible x86 setup?
Anyways, this is all moot anyway: the demand for high density flash is almost entirely in the memory card market. No embedded system realistically needs more than 1GB internal memory (and genera
Re: (Score:2)
It occurred to me just after posting that there would be a decent sized chunk of market in DRMed devices that couldn't allow the use of memory cards, e.g. the iPods. However, there devices are far from being the entirety of the ARM based mobile computers and their switching wouldn't mean anything significant for the market.
Re: (Score:2, Funny)
There, fixed that for ya.
Re: (Score:1)
No embedded system realistically needs more than 1GB internal memory (and generally 256M is plenty). ... There's just no market for some huge flash somehow tightly coupled to a CPU.
Don't you think that 640K should be enough for anyone?
Re:No sir, I don't like it. (Score:5, Informative)
This has nothing to do with Intel's CPU monopoly (which is only really a monopoly in a fairly narrow segment of the CPU market), it has to do with Intel putting a lot of money into process technology. Even when their designs were inferior to AMD's, they remained competitive because they could fab them on a better process and so get higher yields, higher clocks, and lower power consumption for the same chip than AMD.
Intel's SSD products work with anything with a SATA controller, be it ARM, x86, PowerPC, SPARC, or some custom architecture you just wrote to an FPGA. They are not tied to CPUs in any way. NOR flash often is. You quite often get some NOR flash attached to ARM chips for execute-in-place programs, such as the Symbian kernel and apps, freeing up some of main memory.
NAND flash can only be accessed as a block device, so you can't tie it to a CPU at all easily; it has to go through some kind of controller so the OS can pretend that it's a disk. I suppose you could slap a load of DRAM and a separate MMU and DMA controller on it and have something that would look like a big blob of RAM, but the performance characteristics would be horrible to work with.
Re:No sir, I don't like it. (Score:5, Insightful)
Even when their designs were inferior to AMD's, they remained competitive because they could afford to paid off dell, HP, et al.
Fixed.
Re: (Score:2)
Re: (Score:2)
The only time Intel has had an inferior CPU design was the P4 era. And the P4 had _high_ power consumption, in part because the new 90nm process had high leakage currents. They got themselves around that problem by making bigger heatsinks.
Re: (Score:2)
Was NetBurst P4? Anyway, whether or not it was or wasn't, NetBurst was inferior. If Core hadn't come out of the Israeli lab, Intel was within 18 months of wholesale defection of HP, Dell, and IBM to AMD. Would have happened sooner if AMD had trustable manufacturing capacity.
C//
Re: (Score:2, Interesting)
Wintel duopoly mean anything to you?
Inte£ was an abusive monopolist long before most people think. Starting back in the late 80's.
They are always treated with kid gloves but M$oft gets no quarter....
It always makes me chuckle when some Linux noob quotes "M$" but running on an Inte£ cpu/video.
Re: (Score:2)
There are a couple of execute in place NAND memories available now from Samsung and at least one other company that escape