Intel Desperately Struggles To Maintain Relevance Against ARM

Intel, despite Moore’s Law advantages, and serious price-cutting, still has difficulty competing against ARM.“Intel is currently offering its dual-core Atom Z2520 for 7-inch models and Atom 3735G for 8-inch ones, while MediaTek is mainly pushing solutions such as 8382 for the white-box tablet market, the sources said.
Intel’s 7-inch tablet solution is currently priced at about US$20 and the 7.85-inch one around US$27. With a LTE module, which raises the overall cost by about US$20, a 7-inch Intel-based white-box tablet currently has an ex-factory price of about US$50, the sources detailed.
A 3G white-box tablet using MediaTek’s solution has a competitive ex-factory price of US$39.9, the sources added.”
They even have ASUS going exclusively with Intel thanks to huge discounts.

OEMs are paying attention. They will demand similar sharp price-cuts for desktop/notebook PCs and M$’s price will stick out like a sore thumb. GNU/Linux, here we come…

See Intel, MediaTek become favorites in China white-box tablet industry.

About Robert Pogson

I am a retired teacher in Canada. I taught in the subject areas where I have worked for almost forty years: maths, physics, chemistry and computers. I love hunting, fishing, picking berries and mushrooms, too.

This entry was posted in technology and tagged , , , , , , , , . Bookmark the permalink.

13 Responses to Intel Desperately Struggles To Maintain Relevance Against ARM

  1. oiaohm says:

    http://www.eetimes.com/document.asp?doc_id=1319919
    http://www.cadence.com/cadence/newsroom/press_releases/pages/pr.aspx?xml=103012_14nm_test_chip
    Come on really Deaf Spy is completely wrong. Samsung has 14nm arm chips in the market place. Samsung Exynos 6 and later processor that is found inside Samsung devices.

    It also gets more funny if you read my link at the top. Intel themselves is producing batches of Armv8 at 14nm for another arm using company. This is the problem for twits. Arm chips can be produced in anyone foundry including the all important intel. x86 chips can only be produced in restricted foundries. So when every x86 gets mass production at a particular nm so does arm. The question is cost.

    TSMC is always a few generations behind. Once TSMC has a particular scale the bugs are fairly much worked out. They let everyone else go first and have to retool their factories to fix stuff ups.

    The second link covers the fact the test/prototype chip for arm a 14nm was punched out in 2012.

    Deaf Spy only reason we don’t have 14nm in mass as arm processors is lack of factories equip to make 14nm. All the prototype chips and everything else is done.

    Samsung is one of the few parties with currently working 14nm plants. Due to this Samsung gets to charge an arm and a leg for usage of those plants at the moment.

    Mind you intel has tapped out a 10 nm armv8 chip already. Just does not have the means to mass produce it. Reason for its production arm the company wanted to test the design at 10nm and it works fine.

    There is even a benchmark from intel using arm chips produced at 14 and 28 nm showing the 14 doubles its processing speed.

    Samsung is planing to start there 10nm arm chip production at the middle of this year. They have been producing 10nm flash chips since the April 2013. Flash chips you can work around slightly defective ones as you debug the production process.

    Sorry to say 14nm is getting a little long in the tooth.

    The race between intel and samsung is very insane. With samsung mostly winning.

    Basically by the time TSMC gets to 10nm Samsung most likely will be smaller again. In production it not catch intel its catch samsung.

  2. dougman says:

    I find it humorous how the trolls here try and argue that this or that has NOT been done, as if they know.

    Chip manufacturers have at least capable R&D that work 5-10 years in advance of commercial offerings.

    Its been said that the secret military complex has a 50-year jump on technology.

    No one needs large watt hungry CPU’s anymore, that is a thing of the past.

  3. Deaf Spy wrote, “Intel has demoed prototypes now.”

    ARM and TSMC have demoed prototypes now.

    “TSMC has taped out several 20nm HKMG processor designs and is ready for volume production in early 2014. TSMC expects to let customers start designing and taping out 16nm FinFET chips before the end of the year and production 16nm FinFET wafers to enter production a year later in Q1 2015.”

    see TSMC Showing 20nm HKMG and 16nm FinFET Wafers at ARM TechCon

  4. Deaf Spy says:

    Mr. Pogson. You keep speaking in future tense. Intel has demoed prototypes now.

    A week ago you wrote how your wife would, but actually didn’t. This is the same we have here.

  5. Deaf Spy wrote, “Mr. Pogson: b>”ARM can be produced at 14-22nm as well”
    But it is not yet. Not this year. Not even next.”

    30% of TSMC’s wafer revenue was at 28nm in 2013… They began 28nm production in Q4 2011. They have some test production at 20nm already. Moore’s Law suggests they will be at 20nm by 2014 and TSMC is moving faster than Moore’s Law with other technologies as well as resolution improvement. They have 10 and 16nm in their sights. Test production of 20nm ARMed SOC was announced in February, 2014. ARM and TSMC took only 6 months to go from design to test-production. That’s 64-bit multi-core stuff.
    “According to Dr Liu TSMC’s 16FinFET process is now product qualified and leverages their grand alliance program of EDA vendors, customers and IP providers. TSMC will be quickly ramping multiple tape outs across industry segments, and already have more than 20 customer tape outs scheduled for 2014. TSMC’s transistor improvements from 20SoC to 16FinFET can reduce the area by 15% giving continued area scaling as opposed to the FUD from Intel that TSMC’s scaling stalls at the transition point. SemiWiki – TSMC Responds to Intel’s 14nm Density Claim” see FinFET production and the ARM Ecosystem: TSMC readies 16nm FinFET ramp and tips 10nm FinFET plans

  6. Deaf Spy says:

    Mr. Pogson: b>”ARM can be produced at 14-22nm as well”
    But it is not yet. Not this year. Not even next. Intel are on the lead here. Samsung have managed to put out a chip, but are not even close to mass production of 14 nm ARMs. Intel, on the other hand, are looking forward the end of 2014 to start mass shipment.
    You may accuse Intel of imaginary sins as long as you like. There is no proof for that.

    Considering the short life of tablets and smartphones (two-three years maximum), the game is still on. Intel may be late to the party, but they are already in the house, and strong. The fact is that Intel have Atoms superior to any ARM out there for the same power consumption. With full x86 compatibility. And they keep improving the chip.

    Again, Mr. Pogson, you are entitled to your opinion, but not your facts.

  7. Deaf Spy wrote, “Right now, Atom-powered tablets feature the same battery life as ARM-powered ones, but provide higher power and full x86 compatibility.”

    ARM can be produced at 14-22nm as well to further reduce costs. At these levels power-consumption will soon be irrelevant but x86 will still be more expensive to produce. It seems to me Intel is reducing prices on Atom subsidized by the prices of other CPU-lines but there is no point unless Intel can ship units and stem the tide of ARM. That’s not happening. Intel is thinking to ship 40million only Atoms this year for tablets. That’s what, 10%? The horse is out of the barn. The train has left. Even if Intel paid OEMs to use Atoms the white-box guys would still find ARM useful. Then, there are unfair-trading-practice laws… It’s generally not legal to sell below the cost of production. Neither Intel nor M$ can stifle this competition. The last straw is retail space which Wintel lost to Android/Linux.

  8. oiaohm says:

    Deaf Spy I am sorry I would not be saying Intel is more powerful. Next allwinner cpu in the same power usage level is a 8 core 64 bit beast. The battle between arm and x86 is getting very close.

    What one is more power effective is going to be hard.

  9. oiaohm says:

    Robert Pogson there is a bottom point that costs cannot go any lower due to shipping and materials. Modern Atom cpu is not without a GPU the reality the Atom might in fact have lower cpu usage than the Beast due to its hardware mix for video playback due to advances happening with gpu acceleration.

    Even items like libreoffice are using gpu more. So the metric is no longer just CPU either.

    Its going to get interesting. Remember arm is a few cents per cpu above material costs.

  10. Deaf Spy says:

    Desperately? It looks like Intel will be swiping the floor with ARM. Right now, Atom-powered tablets feature the same battery life as ARM-powered ones, but provide higher power and full x86 compatibility. Intel just started putting serious efforts into Atom a couple of years ago. Just watch the show,

  11. oiaohm wrote, “a race to the bottom going on”.

    That’s “oldman”‘s phrase… Using small cheap computers is not a race to the bottom but a sharp increase in efficiency, getting what you want for the least effort. The vast majority of us are consumers and the heaviest use of the CPU is to decode some multimedia. Beast gets up to 10% with that. A smart thingy or an Atom get’s much higher utilization but still not maxed out. e.g. I have an Atom in my home. When folks are watching a DVD movie, I can still SSH in and update the system or transfer files. That’s a dual-core Atom at that.

  12. oiaohm says:

    http://www.bloomberg.com/news/2014-02-22/microsoft-said-to-cut-windows-price-70-to-counter-rivals.html

    The reality is there is going to be a huge volume of hardware under $250 USD. I guess we will see the atom desktop machines back again.

    There is a race to the bottom going on.

Leave a Reply

Your email address will not be published. Required fields are marked *