Slashdot Mirror


ARM Expects 20-Nanometer Processors By Late 2013

angry tapir writes "ARM chips made with an advanced, 20-nanometer manufacturing process could appear in smartphones and tablets by as soon as the end of next year, the head of ARM's processor division said Monday. The more advanced chips should allow device makers to improve the performance of their products without reducing battery life, or offer the same performance with longer battery life."

23 of 73 comments (clear)

  1. next battle? by LavouraArcaica · · Score: 2

    The question is: when Arm will become real Intel competitor? There is already a lot of android laptops, but still lame ones. Sooner or later Arm will become Intel's big problem...

    1. Re:next battle? by fuzzyfuzzyfungus · · Score: 5, Interesting

      I suspect that the answer is a combination of 'don't hold your breath on that' and 'at least a year ago, did you miss it?'

      In terms of sheer screaming power(and, for the moment, even supporting 64 bit memory spaces) ARM is a toy and shows no terribly strong signs of making any strides in that area that Intel would really be worried about.

      On the other hand, it would appear that an awful lot of netbooks and laptops were never sold, possibly never even built, because of tablets and smartphones... If things like this turn out to be a good fit for some 'cloud' niche or other sales of select Xeons could see similar hits.

      At least so far, you don't go up against Chipzilla benchmark-for-benchmark. The world evolves around you such that your virtues are now more desirable than his...

    2. Re:next battle? by GGardner · · Score: 4, Insightful

      ARM based chips will never be real Intel competitor, in the same way that Intel chips never were a real competitor for IBM z-system class mainframes.

    3. Re:next battle? by 0123456 · · Score: 4, Funny

      once apple releases a cheap macbook, watch out dell/hp/asus and others

      Bzzt. You used the words 'Apple' and 'cheap' in the same sentence.

    4. Re:next battle? by hairyfeet · · Score: 4, Insightful

      What I'd like to know is....why does this question keep coming up? Its as stupid as saying "When will mopeds replace Mac trucks?" so the correct answer should be "Well that's just a stupid fucking question!" followed by hitting the moron who asked it with a fish, maybe a nice trout or red snapper.

      While there is a LITTLE overlap, there's not much, and mostly the two are as far apart as mopeds and trucks. The ARM is made to go in little things and is designed for using little power above all, the Intel needs more space but the IPC is just insane so you can get a hell of a lot more work done. You wouldn't try to run Photoshop on your cell would you? So why would you think the chip in your cell would replace the one running Photoshop? ARM simply can't ramp up to the IPC of a P4, much less a Core CPU and if it tried it'd be just as big a user of power as an Intel so it'd be pointless.

      So can't we all just accept that some things are good for some jobs and other things are good for others and leave it at that? Neither chip is gonna "beat" the other because except for a couple of tiny overlaps they really aren't in competition.

      Oh and a final note, the reason that laptops and netbooks aren't selling as much is the same reason desktops aren't selling as much and that is because we went past "good enough" and into "insanely overpowered" and frankly once we got rid of the space heater chips like P4 Mobile that killed the laptops by thermal cycling honestly? with just a tiny bit of TLC they can last a looong time. I know people still using early Core and Turion laptops simply because there wasn't a point in replacing them. After all a new battery was only $30 and there hasn't been any "killer apps" other than games for years and most folks don't game on a laptop.

      So again I wouldn't say one "killed" the other, as everyone I know that has a smartphone or tablet ALSO has a netbook or laptop AND a desktop, its just they don't replace the X86 units until they die because there really isn't a point. While ARM is undergoing its own MHz war frankly for most users even a bottom of the line Pentium or even E350 netbook or laptop is more than enough power for what they are doing with them so they stick with what works. I predict we'll see the same thing with ARM once quads reach mainstream, they'll run into a power wall just as X86 ran into a thermal wall and we'll see the same thing we do now with X86, nobody tossing until the previous one dies.

      --
      ACs don't waste your time replying, your posts are never seen by me.
  2. Hmm... by fuzzyfuzzyfungus · · Score: 5, Interesting

    Does anybody know who the 20nm fabs ARM is expecting to provide these chips are? It was my understanding that Samsung, Globalfoundries and TSMC were still working on a larger process(28mm?) and Intel has been very cagey about fabbing any 3rd-party stuff except for a handful of FPGAs and other high-margin oddballs that don't compete in Intel's area of business in any meaningful way.

    1. Re:Hmm... by pip1 · · Score: 4, Informative

      thats what TSMC is for and why ARM inc have them and IBM etc as core partners for their tape out implementation program as in
      http://www.eetimes.com/electronics-news/4229820/ARM-TSMC-design-20-nm-A15-processor
      from way back in 10/18/2011

      ARM said it would now optimize its physical IP to the TSMC 20-nm process for power, performance and area and produce a specification for a Cortex-A15 processor optimization pack (POP). It did not say how soon this would be completed.

      "This first 20-nm ARM Cortex-A15 tape out paves the way for the next generation of SoC integration and performance," said Mike Inglis, general manager of ARM's processor division, in a statement. These SoCs will be suitable for smartphones, tablet computers, digital home systems, servers and wireless infrastructure, ARM said.

    2. Re:Hmm... by Sycraft-fu · · Score: 4, Interesting

      TSMC is shipping 28nm, latest tech GeForce (GTX680 and 670) and Radeon (7970 and 7950) cards are TSMC 28nm. They do not have anything smaller yet.

      In terms of being behind the curve, yes and no. It is behind Intel but everyone is, always. Intel is generally almost a complete node ahead. Nobody else is doing 22nm, nor will they be for some time. The timeframe ARM is talking about is right along when most companies will start doing 22nm, or its 20nm half node (a number of companies do node and half node processes, some like TSMC are going straight for half nodes).

      So they aren't behind the curve except for Intel, but then everyone is behind Intel. Only Intel is willing to spend the billions in R&D to forge ahead like that and build the fabs at the pace required (their 14nm fab, they are going straight to the half node next time, is going up in Chandler AZ right now).

      I do imagine this is actually directed at Intel though. ARM is getting worried. Intel keeps producing lower and lower end chips, and they are encroaching on ARM's market. Right now it isn't a huge problem particularly since Intel is reserving their latest node for desktop and laptop CPUs. However if Intel starts making 22nm parts to compete with ARM, that could be a problem for ARM. Eve if the Intel parts were less efficient, size can make up a lot of that.

      So they are probably trying to convince partners "stick with us, we'll be there soon!"

    3. Re:Hmm... by pip1 · · Score: 5, Informative

      Guspaz, they skipped 22nm and went directly to 20nm as per their original 2010 plan

      http://www.eetimes.com/electronics-news/4088580/TSMC-skips-22-nm-rolls-20-nm-process
      "Mark LaPedus
      4/13/2010 1:30 PM EDT

      TSMC skips 22-nm, rolls 20-nm process ....
      (TSMC) is putting a new spin on its strategy: After the 28-nm node, it plans to skip the 22-nm ''full node'' and will move directly to the 20-nm ''half node.''

      At its technology conference here, the world's largest silicon foundry also provided details about its 20-nm CMOS process, which will be the company's main technology platform after the 28-nm node. TSMC will also not offer an 18-nm process.

      TSMC's 20-nm process is a 10-level metal technology based on a planar technology. It will feature a high-k/metal gate scheme, strained silicon and newfangled ''low-resistance'' copper ultra-low-k interconnects--or what it calls ''low-r.

      '' For the 20-nm node, it will only offer a high-k/metal-gate scheme for the gate stack--and not a silicon dioxide option.
      TSMC (Hsinchu) will continue to use 193-nm immersion lithography at 20-nm, but it will also deploy a double-patterning and source-mask optimization schemes.

      Unlike its previous processes in recent times--which focused on low power first--TSMC's initial 20-nm process will be a high-performance technology. Following that process, it will roll out a low-power technology.

      With the announcement, TSMC is seeking to gain an edge over its leading-edge rivals, such a GlobalFoundries, Samsung and UMC. ...."

    4. Re:Hmm... by Tough+Love · · Score: 2

      Half node is the new full node.

      --
      When all you have is a hammer, every problem starts to look like a thumb.
    5. Re:Hmm... by Tough+Love · · Score: 2

      ARM is right to be worried; Intel's first production smartphone, despite being single core, was able to produce similar performance and battery life to comparable ARM phones.

      Even if true, it's not enough. Intel also has to ship at a similar price, and given that ARM just takes a few cents per chip in royalties that could present a problem for Intel's margins. They could always try dumping of course but that would be a Sherman act violation.

      --
      When all you have is a hammer, every problem starts to look like a thumb.
    6. Re:Hmm... by bored · · Score: 2

      Its even worse for ARM, because intel was doing it with a binary translation layer, and a older/slower version of android.

      I would be pissing my pants if I were ARM right now. Intel has done this a number of times, some company claims intel can't do X, and a few years later, that company is gone because intel did it.

      The smartphone market is especially problematic because ARM has failed to deliver a proper platform, and therefore vendor lock. So there are dozens of variations that are all incomparable with each other. So the market is flexible, they could just as soon switch to x86, superH, MIPS etc, as the next generation of ARM.

      That said, I don't think ARM really has anything to worry about long term, they will just revert to where they were 8 years ago, when they were in all kinds of embedded products, only now they have more mind share.

  3. wouldn't hold my breath by marcuz · · Score: 5, Interesting

    They have problems delivering 28nm right now so take the 20nm predictions with a pinch of salt. "However, the transition to 28nm does not appear to be going smoothly. ARM heavyweight Qualcomm was the first to introduce a 28nm design, the stunning Snapdragon S4 based on its Krait core. But the outfit is now struggling to meet demand for S4 chips and it is basically becoming a victim of its own success. Other ARM players, such as TI, Nvidia, Samsung and Apple, have yet to introduce a single 28nm part." -- http://www.fudzilla.com/home/item/27414-arm-hopes-to-see-20nm-processors-next-year

    1. Re:wouldn't hold my breath by afidel · · Score: 2

      Yep, NVidia was scheduled to introduce a bunch of new GK108 based parts last month but due to TSMC production issues at 28nm they aren't even able to make enough of the much, much more profitable GK104 based parts to meet demand.

      --
      There are 4 boxes to use in the defense of liberty: soap, ballot, jury, ammo. Use in that order. Starting now.
  4. Radiation resistance? by serbianheretic · · Score: 4, Interesting

    I often wonder, with traces being made smaller all the time, how does this affect radiation resistance? Are we going to hit the point soon where just laying the chip open in the sunlight creates enough of random electron/hole generation, so that the device becomes useless? We already know that chips must be hardened to work in space, how long until this is true for Earth-tied ones? If someone has an answer, it would be interesting to know.

    1. Re:Radiation resistance? by fuzzyfuzzyfungus · · Score: 2

      Smaller is more vulnerable(all else being equal, which it isn't necessarily). Sunlight isn't really an issue in practice(in addition to being alarmingly indestructable, that black epoxy stuff is about as opaque as it looks); but even your big-chunky-classic-single-transistor-in-a-metal-can will show quite readily measurable photo-sensitivity effects if you chop the can open.

    2. Re:Radiation resistance? by vlm · · Score: 2

      That happened with eproms in the 80s. Even industrial lighting outputs enough UV to erase a eprom. Thats why eproms after a certain era in the 80s always had a paper tape over the clear window when not being actively erased.

      Simply laying a late generation eprom out in the sun for a week or so works pretty well. I've done it when my eraser bulb burned out. Obviously this not economical for a commercial dev, but for a dude fooling around in his basement it works pretty well.

      Also this stretches your definition of useless, but there was at least one low current opamp I used that had to be used in the dark because the plastic case was transparent enough that it was enough of a photodiode to screw up some small signal figure (I think it was the input offset current, obviously not the gain or slew rate). Shining a 60 hz fluorescent light at it created 120 hz noise in the output not from EMI but from being a poor photodiode. I don't remember if this was a thermocouple amp or a psuedo-wanna-be-electrometer.

      --
      "Science flies us to the moon. Religion flies us into buildings." - Victor Stenger
    3. Re:Radiation resistance? by 0123456 · · Score: 2

      Google found in their data centers a bit error rate of roughly 300 per gigabyte per month. That's was in 2009, so who knows how that number has scaled in the last 3 years.

      My home file server has 16GB of non-ECC RAM so I find that hard to believe; it gets rebooted every couple of months for kernel upgrades so if it was true I'd expect it to crash or send bad data from the disk cache more often than it's rebooted.

      I saw an interesting study on a mailing list a year or two back where their tests showed that individual DIMMS either had very few errors or lots over their testing period; if you saw more than one or two errors on a DIMM you might as well toss it because you were going to get a lot more. If true and not just due to chance over the test period they used, that would imply that most errors aren't the result of an external source but some kind of manufacturing issue.

  5. Some background by vlm · · Score: 5, Interesting

    Some background you'll never see anywhere else, written by me:

    First of all these funny numbers come from the ITRS. This is not just random numbers, process stages are not a "preferred number system" like resistor values where statistics determines the weird values. Process stage size steps are indirectly determined by physics. ITRS is an industry association of companies who actually make this stuff. Wikipedia has a page for each stage, yes there is a wiki page called "22nm" or something like that. This "20 nm" process is actually a "half step" from the 22nm process. The next "real" step is 16 nm.

    (opinion alert!) Now this is a half step from 22nm to 16nm and is considered a failure. Put your efforts into cheaper higher yield more economic 22 or advance the field to 16, don't screw around halfway at 20. Another interpretation is oxide thicknesses are getting too small at the 22nm process to make anything smaller like 16nm, essentially they're giving up on 16nm because its economically impossible(end opinion alert!) The rest of my post is pretty much factual, as far as I know.

    Another interesting thing about process sizes is this is a half-pitch (essentially a radius) of an array cell. Its dumb and/or marketing to spec half-pitch instead of pitch if you're talking memory. One cell of memory using a "20nm process" is actually 40 nm across. You'll read all kinds of foolishness about how the interconnects are 20 nm across, or a unit memory cell is a 20nm on a side square, or the oxide layer being 20 nm across (which would actually be Fing huge by current standards). Basically almost all size comparisons will just be random crap and no journalist or marketing PR guy ever makes a correct analogy using half pitch, they'll say absolutely anything other than the correct answer, which has made me laugh for decades now.

    Everyone knows everything comes from China. Including semiconductors. Well, actually, no. There's a nice list of plants at wikipedia. You'll see a lot of US addresses. Yes you can probably buy a knock off 555 or 741 from China, but they have almost no small scale plants at all. Pretty much processors come from the USA and a scattering of small time players around the globe. That's interesting. We (USA) make really tiny processors and really giant industrial machinery and not a whole heck of a lot in between. You want a 500000 ton mining dragline? We got it. You want a 22nm processor? We got it. You want a shoe? No we don't make those in this country anymore.
    http://en.wikipedia.org/wiki/List_of_Semiconductor_Fabrication_Plants

    Finally processes are a moving target and different mfgrs and different products are at each stage. There are a couple plants being built for 16 nm process and there are prototypes of "real 16 nm chips" floating around. 22nm process memory was shipping two years ago, 22nm process CPUs are much harder to design. Intel is already shipping 22nm process family CPUs, so AMD gets a golf clap for promising to catch up later this year with something microscopically better. To the best of my knowledge 11nm is not out of the lab yet even for fooling around with.

    And that's about all I know from making some investments in mfgrs since the 80s, some of which worked, some not so good. Not currently investing in this market, but I still keep up with the times and I do a lot of electronics in my basement.

    --
    "Science flies us to the moon. Religion flies us into buildings." - Victor Stenger
    1. Re:Some background by Guspaz · · Score: 4, Informative

      Everyone knows everything comes from China. Including semiconductors. Well, actually, no. There's a nice list of plants at wikipedia. You'll see a lot of US addresses. Yes you can probably buy a knock off 555 or 741 from China, but they have almost no small scale plants at all. Pretty much processors come from the USA and a scattering of small time players around the globe.

      You're pretty much exactly wrong. The US was, as of 2009, in fourth place for semiconductor manufacturing with a 14% share. Ahead were Japan (25%), Taiwan (18%), and Korea (17%). The largest independent semiconductor manufacturer in the world, Taiwan Semiconductor Manufacturing Company, is based in (appropriately enough) Taiwan.

      When I look at all the different chips in my home and where they were made, pretty much only my Intel processor/chipset was made in the US. The rest? Asia.

    2. Re:Some background by Sycraft-fu · · Score: 3, Interesting

      Ya I think the half-node obsession came from companies trying to "beat" Intel. For quite some time Intel has been almost a node ahead of everyone. Everyone else is ramping up one node and not long after Intel has the next online. So they started doing the half-node stuff. TSMC first did this for 55nm, though they were still behind Intel at the time, but then for 40nm. So they could claim on paper at least to be ahead of Intel. Intel had a 45nm process, TSMC has a 40nm process. Of course not too long after Intel went to 32nm, but for awhile TSMC could claim to be ahead (as you noted it isn't quite so simple).

      Apparently Intel is sick of this and they are going to 14nm next with their new plant in Chandler. How much of that is pure marketing (I could see a hybrid process where most of it is done 16nm but something is done using 14nm half-node process so that you can technically call it 14nm, maybe the cache) I don't know but there you go.

  6. Don't forget Taiwan and South Korea by tempmpi · · Score: 3, Informative

    TSMC, UMC and Samsung are some of the biggest players in the field. And almost all the DRAM is manufactured in South Korea.

    --
    Jan
  7. Re:So ARM are investing in multi-billion chips fab by Guspaz · · Score: 3, Informative

    I'm not sure you understand how ARM's business model works. They don't manufacture chips themselves, and they don't even hire somebody else to manufacture chips for them. They also don't design chips for a specific process node. They just produce a design and leave it up to a company like Texas Instruments to figure out how to build them at a certain process node (or hire some fab company to do it).

    The 20nm statement is just a prediction. They're saying they expect their customers to get 20nm parts out in 2013.