Become a fan of Slashdot on Facebook

 


Forgot your password?
Close
typodupeerror
×
IBM Science Technology

IBM Claims World's Smallest SRAM Memory Cell 206

Posted by CmdrTaco from the less-is-more dept.
nokiator writes "IBM issued a press release today claiming that it has built an SRAM memory cell that is ten times smaller than those currently available. My interpretation of the PRese in this release is that IBM will be able to build 256Mb or 512Mb SRAM chips or integrate 32MB or more SRAM into processor dies for cache applications in the future. Of course, showing some SRAM cell prototypes is a long ways from being able to manufacture this technology in a cost effective way. There is no information in this PR about the speed or power consumption of SRAM blocks that can be built with this new cell technology. This is not likely to be a potential DRAM replacement for mainstream applications as DRAM already offers more than ten times density compared to SRAM at much better cost."
This discussion has been archived. No new comments can be posted.

IBM Claims World's Smallest SRAM Memory Cell More

IBM Claims World's Smallest SRAM Memory Cell

Comments Filter:

  • These memory cells are so small... (Score:5, Funny)

    by Anonymous Coward on Monday December 06, 2004 @03:59PM (#11010600)
    ...they can only store zeros. They are working feverishly on scaling them up to store ones.

  • IBM Rocks (Score:4, Insightful)

    by vmcto ( 833771 ) * on Monday December 06, 2004 @04:00PM (#11010606) Homepage Journal
    And people keep asking me why we don't buy servers from Dell. Maybe I'm old-fashioned in thinking that a company that continuously innovates the technology and not the distribution or marketing channel is whom I should get my technology from...

    • Re:IBM Rocks (Score:1, Funny)

      by scaaven ( 783465 )
      What are you talking about? Dell makes neat cases.
    • Mod parent UP!!! Dell is repackaged Chinese crap....they only make money on the float from their rebate......

    • Re:IBM Rocks (Score:5, Informative)

      by Martin Blank ( 154261 ) on Monday December 06, 2004 @04:15PM (#11010761) Homepage Journal
      Their server division buys most if not all of its parts from other companies, excepting perhaps the PPC chips. Cases, CPUs, memory, video chips, and most likely even motherboards are manufactured by other companies, who probably also have a more direct hand in design than does IBM, which may only do oversight engineering such as reviewing final designs to ensure there are no significant bottlenecks or thermal build points.
      • IBM makes a lot of its own Power and PPC chips, I imagine they make their own chipsets too, although I suppose they could use Motorola chipsets.

      • Re:IBM Rocks (Score:5, Informative)

        by curious.corn ( 167387 ) on Monday December 06, 2004 @04:32PM (#11010913)
        Bullshit... I once saw a low end intel IBM server... yeah... IBM is just like any other incorporated beige box assembler... Dell? Oh, you're so much wrong... listen: redundant monster fan, redundant power supply & cabling, redundant motherboard chipset, HotSwap PCI cards, HotSwap and (configurable) Redundant RAM (have you ever heard of RAM RAID? is that Redundant Array of Expensive Ram?)! You can pop the Bleedin' memory sticks from the machine Live! (I mean, the Dimm socket braces have leds identifying the faulty stick!)... drums roll... HOTSWAPPABLE BLOODY CPU BASKET (two at the same time... the thing waits brainless for the tech to drop the new one)! Yeah... just like that Dell... I'm daed shure this beast doesn't even let Winders touch the metal, there must be some virtualizing layer because I can't possibly believe Windows could ever survive in such an advanced machine without self-formatting in shame! Now boy, go blow your nose and play with your cold-cathode modded water-hyped P4EE...
        • And explain: What are you saying you like, Dell, or IBM? Because I can't figure it out.
          • You're right... I did get into a bumbling ram but it happens; I'm a bit burnt out!;-) Anyway, the rig I saw was (obvously) an IBM and I was enumerating, confusingly I agree, the specs that the tech had gloatingly told me while I was wiping the tears off my face when they assembled it;-) Now, the parent post was playing down IBM like "not much different from Dell" so I thought about that machine... ... ah, BTW, the machine itself was redundant... there were two of 'em... ("Oh my God, it's full of stars...")
            • while I was wiping the tears off my face

              Were those tears of pain, relief, joy or just plain old exhaustion?
            • Okay, so it was an IBM box. I was going to say, there ain't no dell like that that I know of. I have always liked the IBM server hardware.

              And your parent poster was right; IBM actually develops new technology. Although many of their products are created with the same parts as anyone else, they use that revenue to do things like invent new microprocessors, memory, hard drive technologies, you name it. IBM is a cool company, even though they're really big. I feel as though IBM has learned it's les

      • Re:IBM Rocks (Score:3, Insightful)

        by trixillion ( 66374 )
        Even if what you say is true it is irrelevent to the grandparents post. IBM will keep the profit and therefore by buying IBM you are investing in pioneering work.
    • IBM makes quite a few breakthoughs in the lab but they never seem to generate much successful product. Sure they make (or made) RAM chips, microprocessors etc, but they soon get overtaken by other manufacturers. I doubt they ever make money on the silicon they sell. Perhaps they make money when they license their patents etc.
  • They say a combination involving E-beam. That do not smell like mass production.
    • I hate to reply to my own post but a back of the envelope calculation makes that 280x280nm. Cool even if this is a little beyond practical.

    • The problem with E-beam is that it is a serial process. In effect you have to draw every single transistor using a megnetic field to move the electron beam.

      What makes IC manufacturing in general manufacturable is the fact that you are using photo lithography with an optical mask through which you expose the wafer. This is a parallel process whereas E-beam, which can be good for engineering samples, sucks for manufacturability.

  • IBM will be able to build 256Mb or 512Mb SRAM chips

    One thing I've never really known is what does this type of figure translate to in terms of real amount of RAM on a memory module that I would stick in my computer?

    Its always hard to tell because there is such a long time between companies saying that they have made X Mb fit on a chip to the time that they make a 512MB dimm.

    • Memory manufacturers, in order to have bigger numbers, specify capacity in megabits, not megabytes. So, take the number of megabits, and divide by 8. That gives you megabytes. Now, multiply by the number of chips you want to stick on the DIMM (I've seen as low as 2, and as high as 32), and you have the total capacity.

      steve

      • Memory manufacturers use megabits because they have done so for a long, long time. At one time, memory parts with single bit data busses were common. Even now, some devices have data widths other that eight, and not every CPU addresses eight bit bytes.

    • Re:Chips = what? (Score:5, Informative)

      by ottffssent ( 18387 ) on Monday December 06, 2004 @04:09PM (#11010691)
      The memory modules you put in your computer are composed of DRAM chips. DRAM uses a capacitor and a transistor per cell (plus sense amps, decoders, etc.). DRAM requires refresh (the charge on the cap leaks off) but is relatively low-power and very dense. SRAM uses no capacitors, but more transistors (4 or 6) per bit; it's higher-power but faster and doesn't require refresh.

      So, SRAM density has nothing to do with DIMMs you put in a computer. It's used for on-chip caches (and off-chip caches), but is too expensive for main memory. Denser SRAM means that Opteron you've got with 1M L2 cache could have 4M or 8M if IBM can mass-produce the stuff.

      DIMMs usually have 16 chips (18 for ECC modules). So, if you have 512Mbit DRAMs, you put 16 of them on a module and you get 8Gbit = 1Gbyte. Gigabit DRAMs can make 2GB DIMMs. 2Gbit DRAMs are needed to make 4GB DIMMs; they cost hundreds of dollars each (and you need 16!), which is why 4GB DIMMs are so amazingly expensive.
      • Well that answers the question in relation to this article. But generally whenever there is some story in a tech magazine about a big corporation that just fit X Mbits (and I know the relation between a Mb and MB) on a chip, it doesn't quite make sense with real world products.

        For instance, I've seen some articles for RAM talk about companies fiting 256Mbits on a chip, but at the time of the article, 1GB dimms where commonly available and so this article wasn't really a bit deal. This led me to think tha
        • Comment removed based on user account deletion
          • Did you even read what I wrote?
            I understand the difference between a RAM and the chips on it. But what I don't understand is if there is somehow a difference between the chips that they so often talk about in articles like this one and the chips on a ram stick. Because from the numbers that are stated in various articles, I would think that there is a difference.
            Besides, 256Mb X 4 != 1GB Unless you meant to say 256MB.
      • And, to add just a little to the parent, more cache is one of the best ways to improve your performance.

        If you're working on a 2 Meg file and you only have a 256K cache, then your CPU has to work on a little bit of it, then swap the cache to main memory to work on some more, then again, then again. Each of those swaps takes time, and main memory is way slower than cache. So if you can store most of your work in cache and save the trips to memory, your get much faster speed.

        So a webserver with a 2 GHz proc
      • 1) I don't know what I'm talking about.

        2) I thought that SRAM had to be "refreshed" on most every clock cycle, and they store information by refreshing themselves with their current state.

        • Re:Chips = what? (Score:3, Informative)

          by curious.corn ( 167387 )
          Nope, think a pendulum locked at the maximum swing position. Ok, I'm being cerebral so think a stat ram cell like two inverters interlocked so that the output of one forces the other in a state whose output forces the first one into the same state it was before like an oscillator that dosen't oscillate (it would if the second's output would force the first inverter to change state rather than maintain the current).

          It's like a flipflop but wired differently. It's wicked fast because you don't have to refres
    • 512Mb = 64MB assuming an 8b word-size. If you use 32b word size 512Mb = 16MB.
      • You're confusing word-size and byte. For reference, see byte [wikipedia.org] and word size [wikipedia.org].

        Consider: a 32-bit processor has a word-size of 32 bits, and a 64-bit processor has a word size of 64 bits. Typically this refers to the size of the int datatype on the system. But in both the x86 and x86-64 archs, for example, a byte is still 8 bits. So unless you're talking about some old-ass architecture from back in the day, a byte is 8 bits.

    • Since you have about a dozen replies, and nobody has actually ANSWERED your question, here goes:

      Memory chips are delineated in terms of Megabits. Your typical DIMM has 4, 8 or 16 chips...any more requires fun stacking tricks or a double-height PCB.

      The 512 Megabit chip noted here would be 64 Megabytes per chip (64 * 8 = 512). Thus, DIMMs would have the following:

      4 chips = 256MB
      8 chips = 512MB
      16 chips = 1024MB

      Now, as it has been noted, SRAM and DRAM are not compatible, so you'll never see this on a modu
    • Well, for people who use these sorts of chips, this data can be useful. SRAM chips aren't used for main system memory, but is useful as cache. While you could make some computer with only SRAM, nobody would consider doing it, due to the expense. However, knowing this is useful for any chip makers, who might want to figure out how much area to budget out for. So you could figure that you want 256 Cache lines, two-way associative, with 16 32-bit words per line. That's going to be about 128kb, and the con
      • Historical note: While SRAM isn't used for main system memory any more, back in the 1980s it occasionally was. The IBM PC Convertable (1986) and the Apple Macintosh Portable (1989) are the most prominent examples.

  • Density vs Speed (Score:3, Interesting)

    by fembots ( 753724 ) on Monday December 06, 2004 @04:02PM (#11010630) Homepage
    This is not likely to be a potential DRAM replacement for mainstream applications as DRAM already offers more than ten times density compared to SRAM at much better cost."

    I thought the PR implied "Although not as dense, SRAM is many times faster than dynamic random access memory (DRAM).", density is like a also-run.

    • Re:Density vs Speed (Score:2, Informative)

      by gazuga ( 128955 )
      True -- SRAM has a much faster access time and is usually used in caches (such as on-die CPU L1 and L2 caches). However, since it is so much more expensive than DRAM, you don't see it used for system memory.

      • In the PC world, you are correct. SRAM isn't used for main memory. In a lot of embedded applications, though, using SRAM for main memory is pretty common.

    • I thought the PR implied "Although not as dense, SRAM is many times faster than dynamic random access memory (DRAM).", density is like a also-run

      I remmber back in the days of the release of the original Macintosh II, a lot of articles (In Byte, MacUser, etc.) about the new 68020 architecture stated that main memory in the Mac would eventually be transitioned to SRAM because of SRAM's speed and power-consumption advantages. Cheap and dense SRAM was coming, "real soon", so that extra wait states or caches wo

  • They say 50.000 at the end of a human hair. Do anybody know the actual size of this cell?
  • While I think this is great news, I just wonder how much longer it will be until a physical size problems emerges. I know that crosstalk and heat affect current processor designs, so do these issues also occur with memory as well?
    • Crosstalk and heat problems affect anything with sufficiently small circuits, RAM included. Crosstalk is caused by currents in wires inducing currents in their nearby neighbors, and heat is caused by resistance in the wires. The registers in a CPU are a small kind of memory, and they are built using the same technology the CPU is made with, so you can get a feel for how small we can make RAM at our current technology by observing how much we can pack into a given size CPU.

  • Cost effective? That depends... (Score:5, Insightful)

    by cavac ( 640390 ) on Monday December 06, 2004 @04:11PM (#11010726) Homepage
    Of course, showing some SRAM cell prototypes is a long ways from being able to manufacture this technology in a cost effective way.

    Well, were still talking about IBM here? Do you really think that a few hundred dollars more would even get noticed at clients that buy a server in the 100K range?

    The main advantage of buying high-end gear from IBM, Cisco and the like isn't that you get cheap hardware ('cause you simply don't). You buy the gear from that company because you get 10 years in-house service including remote failure detection if you pay for it. That means, THEY call YOU before you even notice one of your tripple-redunant drives has problems. At this point in time, the technician is probably already on the way up to your office.

    Sure, it's very expensive. But you save quite a lot by not having any significant downtime...

    Seen in that context, 500 bucks more for RAM is IMHO just irrelevant to even think of...
    • That's a minute fraction of the total market for RAM. As much as some people like to believe it, the world does not revolve around large business. Large business is actually somewhat of a niche market.
      • Yes and no. But the large business is actually what drive the low-end market.

        Generally, if you open business and buy a branded servers, you're stuck to that brand a long time. Naturally many companies look for manufacturers that also offer the real big-league stuff, so if they ever need a big iron they don't have to switch the manufacturer.

        AND, for selling a real big super-computer server farm (Top1000), which makes real, hard millions of profit, you have to have the big irons in stock.
        • AND, for selling a real big super-computer server farm (Top1000), which makes real, hard millions of profit, you have to have the big irons in stock

          The only problem is that you need BILLIONS, not millions to develop the next memory generation....
    • If it works, I think it stands to benefit anyone that buys a computer based on this technology. I think IBM licences a lot of their technology. Because of their research, their patent portfolio is very large and I think growing by dozens if not hundreds per month.

      If they manage to licence this out, it can pay back that research quite handsomely, and we benefit when it is incorporated into the Pentium, Athlon and PPC chips if it increases cache capacity, maybe bandwidth & latency too, while still incr

  • SRAM has plusses and minuses. (Score:5, Funny)

    by Sheetrock ( 152993 ) on Monday December 06, 2004 @04:11PM (#11010730) Homepage Journal
    As mentioned in the article, SRAM (Static RAM) is many times faster than DRAM (Dynamic RAM) while simultaneously offering a smaller footprint.

    However, I wonder if the additional implementation requirements justify the benefits. Static typing is only found in certain computer languages, and programmers have come to rely on dynamic memory allocation offered by malloc() or similar routines. I suspect with careful design one could fully exploit the advantages present -- with software being cheaper than hardware, it could easily be well worth it in embedded or pre-fabricated devices.

    The type of implementor that uses (dynamic) extreme programming methodologies may be left out in the cold, although I would like to suggest that would occur anyway to a person working without a blueprint. Regardless, it will be exciting to see how this develops from the embedded perspective...

  • This is *not* true! (Score:2, Interesting)

    by Anonymous Coward
    Intel released a press release [intel.com] four years ago claiming to have built SRAM memory cells which (to my calculation) work out about three times as small as IBM's. IBM are clearly claiming to have invented the smallest commercially viable memory cells, but not the smallest outside of that context. Maybe /. should try to keep up a bit more.
  • Every few months or so, new smaller stuff always comes out. I never see much that takes advantage of the stuff, though. That's unfortunate. Maybe it's just used in non-popular, expensive products for big corporations.

    I guess this kind of stuff won't matter to the average user for another 10 years, if ever.

  • The Real Cost of Dram (Score:4, Interesting)

    by Nom du Keyboard ( 633989 ) on Monday December 06, 2004 @04:26PM (#11010854)
    DRAM already offers more than ten times density compared to SRAM at much better cost.

    Excuse me, but isn't the cost of any feature directly related to its size, making the above statement self-redundant?

    I mean, a wafer-start is a fixed cost, divided by the number of processors it yields. That makes the area of the processor die directly relate to its cost, and the size of any feature relates to its subcost portion of the overall processor cost.

    Or in simple terms: Smaller features should always cost less because you get more of them per fixed cost wafer.

    Am I missing something major here?

    • All manufacturing processes are not the same; in particular DRAM and logic (SRAM) use different processes.
    • Well, you've basically got it.

      The only other thing to think about is the yield of the wafer (ie. how many good dies you get from each wafer).

      Some circuits are harder to build or more prone to failures in the manufacturing process - ADCs, DACs and high-speed opamps come to mind. Some of the flaws are inherent in the way the wafer was grown.

  • If you can shrink SRAM down to a size equivalent to DRAM, then it CAN serve as an effective replacement, and here is why:

    1) No fancy control logic. DRAM needs to be refreshed on a regular basis. SRAM is a straight "chip select, read/write" type of ram.
    2) low power. Because it is not being constantly refreshed, it can hold those bits with far less power. Thats why you see NVSRAM and don't see NVDRAM. Imagine having 1 gig of RAM that is battery backed up?
    3) One can argue that without the control logic, it wi
    • All of this is true, of course, but even given the right circumstances it won't replace DRAM any time soon, simply because it involves a rather significant change in memory controller design (which, on modern systems, are beggining to be integrated with the CPU itself).

      Still, i HOPE SRAM becomes as pratical as DRAM for everyday computing some day.
    • Actually SRAM typically requires _MORE_ power than DRAM because SRAM consumes power constantly where DRAM's only really consumes power during refresh cycles.

      So point 2 is wrong, but points 1 and 3 are right. (actually, not just theoretically faster, _DEFINITELY_ faster)

    • I was under the impression that even though SRAM doesn't require refreshing, it takes more power. If you're designing for speed (and low noise) you'll have 6 transistors in your SRAM cell (as oppose to 1 for DRAM). Granted, these are all tiny transistors (with a set of very large transistors for amplification), but so are the DRAM transistors.

      The slowest part of ram is the row (column?) select. (I forgot which-- one is around an order of magnitude slower I believe) I don't know much about the bleeding
      • As far as I know SRAM is gates and DRAM is capacitors and therein you will find the issues involved with their manufacture. Unfortunately I know shit about this stuff, I can't even remember the basic theory like what's doped with what and even that was only old CMOS stuff that I learned and won't even apply now. Anyway moving on IBM is supposed to be working on this MRAM stuff as well, some sort of electrolitho-manufactured magnetic bubble core memory or something. From what I understand the magnetic fields
        • I saw a presentation on MRAM less than a month ago. Anyway I forgot it already. Right now it's slower than DRAM, however not much slower. The idea is it's very new so refinement could easily push it past DRAM in terms of speed and density. There was some aspect of it which did take a lot more power (writing I believe?), but I think it was offset by the rest of the operations (read and idle) being lower power.

          They're looking into it for "instant-on" technology (say, store a few megs of bootup code you'
    • regarding the power consumption, I made reference to the power required to hold the bits, not read or write. I am pretty sure that, while SRAM requires more power to read or write, holding those bits requires little power.
    • Comment removed based on user account deletion

Slashdot Top Deals

"Conversion, fastidious Goddess, loves blood better than brick, and feasts most subtly on the human will." -- Virginia Woolf, "Mrs. Dalloway"

Close