Hacker News Reader

They don't need Atom, as far as mobile devices go they are bringing Core to ARM's power envelope. You can get Core M CPU's today with 3W power envelope with considerable performance lead over most ARM SOC's.

Intel has made a bet that it would take just as much time for ARM to reach x86 level of performance as it would take Intel to bring x86 power consumption down to SOC levels. Now Intel can play on both sides with their low power x86 parts, Xeon-D with upto 16 cores and Core-M in 2c/4t configuration.

That core M will be $1xx compared to $1x-2x for ARM chips.

That's economy of scale I don't see a reason for it to be more expensive the Apple A8x had a higher transistor count than an 8 core Core i7 Haswell-E CPU. Also I'm pretty sure that high end SOC's that end up in iPhones, Galaxy S's and the like cost more like the Core-M than the 20$ or so you would pay for a MediaTek SOC.

mediatek socs are actually $4(this is a price from 2014, when Intel had to match that $4 price while subsidizing Atoms), and that includes pmic. $20 is a price for a hieng quad core chips.

I've design with Intel Atoms and ARM chips. Atoms are very bloated and require a lot of external support to get up and running. Intel's firmware/boot support is atrocious. They still have PC OEM's in mind. ARM chips on the other hand are much easier/cheaper to design for. The SoC is better integrated and targeted towards a low overall BOM cost. Core M is worse than Atom.

Not true at all. If Intel could do that:

1) they wouldn't have replaced its Core-based Celerons and Pentiums with Atoms to increase profits - and they sell those for $110-$160

2) they wouldn't have licensed Atom IP to Rockchip and other low-end chip makers, essentially pursuing an ARM-like IP-revenue model (for which they would've made peanuts and didn't pan out anyway).

Intel Celerons even the embedded ones are currently based on Skylake. https://en.wikipedia.org/wiki/List_of_Intel_Celeron_micropro...

Intel licenses allot of things they still have ARM licenses and probably even Power PC ones.

The recommended customer prices for older Core M's is listed as $281

>Now Intel can play on both sides with their low power x86 parts, Xeon-D with upto 16 cores and Core-M in 2c/4t configuration.

There's only one little problem: the market has settled on ARM and doesn't care anymore.

So how are they gonna play those "both sides" again?

Actually the market for mobile yes. But on servers it will actually take a long time to replace ARM. So it's actually a win for customers when these two are rivals since we get cheaper servers or lower tdp on servers and cheaper mobile devices.

Server side, there is an initiative to get rid of Intel as well[0][1] with a PowerPC based chip. It'll be interesting to see if it ends up working, but I think everyone wants to get away from Intel's high prices.

[0] https://cloudplatform.googleblog.com/2016/04/Google-and-Rack...

[1] https://en.wikipedia.org/wiki/OpenPOWER_Foundation

Not really, the market has settled on CPU-independent bytecode. I've got a x86 tablet, and you don't notice it at all, other than the Intel logo on the back. As long as there is an ART port it could be MIPS and still have pretty much the same compatibility.

While this is mostly true, it is just untrue enough to cause trouble for a non-negligible number of apps.

The Intel Houdini software used to make Android on X86 possible is a marvel of engineering but it is not perfect. I worked for a company that made an Android tablet around an Intel chip instead of ARM. We were hit with a great number of user reports of certain apps crashing, overheating, or simply refusing to run in the first place. We verified in QA that these apps would work just fine on ARM tablets with the same Android version and similar specs.

To be fair to Intel, they were very good to work with and really did (and do) want to make it work. I have no doubt they will. Eventually.

... Compiler.

Yes, all major mobile OSes have followed mainframes and have settled in CPU-independent bytecode, for most part, as long as C or C++ code isn't used.

Currently iOS is the only one that also supports having C and C++ as bytecode (LLVM one).

Android and WP, only support it for Java, RenderScript and .NET languages.

AFAIK if app uses NDK it has to have compiled code for every architecture. If developer didn't make version for x86 you won't run it.

Single core Atom (2012 Z2480 on Razr I) at 2GHz has significant perceived performance advantage even compared to processors released after. It's way snappier than the Moto G quadcore ARM cpu from 2013. I don't know how much battery life is sacrificed for this though. But I'd be very eager to try a Core M smartphone, since Atom has always been the low-end of Intel cpu architecture. That said I lost track of Intel marketing, maybe Core M is just a rebranding of the same or similar arch.

There's a very large difference between a Cortex-A7 in a Moto G and a Cortex-A57 or Cortex a72. The latter are about 3x faster.

Atom has barely increased its single-threaded performance from like 300 pts in Passmark to about 540 right now (in Pentiums, in PCs, which cost $160 a pop).

Intel is not competitive on price. That's why Atom didn't pan out in mobile. They tried to do it by heavily subsidizing its high-end mobile chips, which typically cost close to $50, to compete against a $15 mid-range Qualcomm chip. That's why some people were "impressed" by what an Asus Zenfone could do for instance, for its mid-range price overall.

And if Intel can't compete with Atom on price, then there's NO CHANCE it will ever compete with Core M in mobile. The only reason it even exists in so called "tablets" that go for $1000 right now, is because Microsoft failed to make a good case for ARM-based Windows machines with its poor app support. But that will always remain a niche.

> And if Intel can't compete with Atom on price, then there's NO CHANCE it will ever compete with Core M in mobile. The only reason it even exists in so called "tablets" that go for $1000 right now, is because Microsoft failed to make a good case for ARM-based Windows machines with its poor app support. But that will always remain a niche.

And outside the Surface those aren't selling well either.

(IMO rightfully so, the Venue 11 Pro is the worst device I've seen in years – hypothetical hardware dickwaving aside, even a $50 OEM Android tablet has better UX and usability than Windows 10 without mouse/keyboard.)

Asus T-100 TA series work pretty well as Windows 8/10 tablets.

Yes, because they ship a keyboard and touchpad. But you wouldn't buy the Asus ZenPad instead and try to use Windows 8/10 solely with touchscreen, would you?

It's the complete opposite of how iOS/Android tablets operate.

That moto G had great battery life I could tell you.

Core M is the "same" CPU as their server and desktop CPU's. It's not based on the Atom line. https://en.wikipedia.org/wiki/Skylake_(microarchitecture)

First off, those "3W Core CPUs" are a scam and a dud. They will overheat and throttle.

Second, they cost WAYYYY more than a mobile chip. They cost more than an iPhone's ENTIRE BOM, or about 10x more than what Apple pays for its chips.

> First off, those "3W Core CPUs" are a scam and a dud. They will overheat and throttle.

Throttling has been a problem on high-end Androids for years now.

> Second, they cost WAYYYY more than a mobile chip.

That's the real issue – the margins in the mobile market are ridiculously slim, and there is no x86 monopoly Intel can abuse to lock in vendors and customers. So cheap ARMs it is.

so intel is going to start selling CORE processors for 1-5usd apiece?

> why Intel gave up on XScale

The gist of the Digital Semiconductor acquisition was that DEC had kicked Intel's ass in circuit design but turned out to have insufficient volume to be profitable, so Intel and DEC agreed to move DEC onto the upcoming Itanium and DEC could get rid of their bleeding edge but money losing semi business. Alphas and XScales were both remarkable feats - on high performance and low power end, respectively. DEC agreed to kill Alpha and move to Intel's Itanium - but Intel had no existing ARM style product at the time.

Intel was essentially forced into buying XScale from Digital via a legal settlement. DEC sued Intel for uncompetitive behavior (and looked likely to win).

I don't think Intel really wanted ARM and so made half-hearted attempts at selling it.

The Alpha's extremely weak memory order gurantees would have doomed it in the multicore market. Its hard enough to get multi threaded code correct on X86.

This is meta, but you can make the quoted text more readable if you add a newline after some X number of words, like this:

    Otellini said Intel passed on the opportunity to supply
    Apple because the economics did not make sense at the
    time given the forecast product cost and expected
    volume. He told The Atlantic, "The thing you have
    to remember is that this was before the iPhone was
    introduced and no one knew what the iPhone would do...
    At the end of the day, there was a chip that they were
    interested in that they wanted to pay a certain price
    for and not a nickel more and that price was below our
    forecasted cost. I couldn't see it.

    "It wasn't one of these things you can make up on
    volume. And in hindsight, the forecasted cost was wrong
    and the volume was 100x what anyone thought."

Oh, so that's how it's done? I always thought there was a formatting rule I was missing. Thanks.

It could be done automatically as it is indeed confusing. Each new line still needs 4 spaces though... I think!

One downside of this is that it becomes very unreadable on small devices (the user has to horizontally scroll back and forth to read each line, ugh)

Representing quoted text in italics does not suffer that downside.

Oh, I never noticed this as I use an app to read HN (forgot the name).

I was an industry analyst at the time and I think the short answer is that Intel made the decision to pursue an x86-everywhere strategy. They thought 1.) That it would benefit them if there was a the same x86 architecture from server to mobile and 2.) They thought Atom would meet mobile requirements.

(1) was doubtless true. But (2) didn't pan out.

One can find slides from IDFs etc. at the time with Intel trying to buttress its x86-everywhere arguments by showing that Flash ran more consistently on small devices when they were x86-based.

In retrospect, Atom was mostly wish fulfillment. I can't say whether this was an execution issue or whether the strategy was intrinsically doomed to fail.

Anything in Intel that's not x86 gets killed by internal politics. Somehow Intel didn't understand that when they bought it.

Eh, don’t forget the external forces. Remember Itanium vs AMD’s x86_64?

Itanium died because it was a very niche architecture Intel was working on their desktop/common server x64 line when AMD came out with Athlon64 which more or less won because it's desktop performance over the Pentium 4. Itanium was a RISC-y endeavour by Intel and it had it's problems but you can't really say it died because of x86_64.

Itanium was a RISC-y endeavour by Intel and it had it's problems but you can't really say it died because of x86_64. Microsoft adopted it and dropped IA-64 like a hot potato. Intel dropped Yamhill and other projects that attempted to bring IA-64 to the desktop as soon as they realized MS wanted x86-64. Performance didn't factor in because AMD would have never had the chance to provide the world all the CPUs it wanted (supply constraints)

I think you are mixing a few things here Yamhill was supposed to be a x86_64 CPU (under license from AMD) it was dropped in favor of promoting Intels own IA64 instead. http://www.geek.com/chips/intels-otellini-says-no-to-yamhill...

Microsoft dropped support for IA64 only a few years ago with Windows Server 2008 R2 being the last OS that supported it due to the limited market share. This was quite long after Intel has killed IA64 internally on it's own.

Itanium was a trojan horse to get their rivals to give up on MIPS/Alpha/UltraSparc development. As soon as those CPUs lost all traction, Intel dumped Itanium as well. Intel was never serious about the Itanium. Considering how performance sensitive it ought to be, Intel chose to build Itaniums on older manufacturing processes. The fact that Itanium performance was hopeless should not have been a surprise.

Intel chose to build Itaniums on older manufacturing processes

They didn't have any choice. Itaniums were big, and their defect rate on newer processes was too high to get any yield on chips that size.

Itanium was EPIC, pretty much the opposite of RISC. Intel bet on compilers being good enough to tell the CPU which branches it could execute in parallel, and lost that bet.

Er... those are orthogonal concepts. Itanium had a simple instruction set with primarily register-register operations. The explicit ILP part has nothing to do with that.

Perfect example with their best replacement for x86: the i960. It was one of most impressive compromises I've seen from the time period in terms of speed, reliability, and security.

https://en.wikipedia.org/wiki/Intel_i960

Some variants are still in production mostly for legacy purposes but not the one (MC) I wanted. Too bad.

http://www.intel.com/design/i960/index.htm

Why have you posted that link?!?

Now I will have to research about BiiN and Intel iAPX 432. :)

Start with the classic!

Performance Effects of Architectural Complexity in the Intel 432

https://www.princeton.edu/~rblee/ELE572Papers/Fall04Readings...

Thanks!

Yep, yep, the other important one. Of course, one could read this along with the i960 and System/38 papers to see if any good ideas pop up.

I posted it a bunch. I usually post it, Burroughs B5000, and capability-systems (includes AS/400 predecessor) links [1] together. Intel's i432 was an amazing attempt to clean-slate the machine of future. Safe, manageable, and consistent from the ground up. Just overdid it in terms of hardware requirements and certain components should've been firmware for easier improvements.

The i960/BiiN system improved on that by greatly reducing complexity of the hardware. It was a fast RISC system, had all kinds of error detection, supported HA configuration, and had 432's object-descriptor protections. Object- and page-oriented system. I expect our computer security and reliability situation on Wintel might have turned out differently given what highly-secure systems did with x86's shitty segment/paging/ring combo and HA with lock-step.

Note: A Slashdot article on legacy systems once knocked the F-35 for using old i960 CPU's. Thought that was stupid idea. I think designer was probably thinking too many steps ahead with the market ruining it for him/her.

[1] http://www.cs.washington.edu/homes/levy/capabook/index.html

[2] http://craigchamberlain.com/library/blackhat-2007/Moyer/Extr...

My take is Intel doesn't really understand markets where they need to fight for market share. They invest in something, it doesn't become a large profit business in a few years so they kill it. Customers see this and passively/actively avoid Intel products when possible.

Passive avoidance: Customer doesn't even think to consider what products Intel is offering. Active avoidance: Tries not to spec Intel products when possible.

The internal politics part comes from this scenario. Say someone makes a good proposal, spend $20 million. In return a get product line worth 100 million yr gross and $40 million net. Manager thinks, $100 million a year isn't going to get me a VP position at a $50 billion dollar company.

> Say someone makes a good proposal, spend $20 million. In return a get product line worth 100 million yr gross and $40 million net. Manager thinks, $100 million a year isn't going to get me a VP position at a $50 billion dollar company.

You could also be describing execs at Apple.

Friend of mine worked at Apple in the period between Jobs 1.0 and Jobs 2.0. He described an environment of management 'wolf packs' slowly destroying everything. Stuff like this, guy gets promoted as a manger of a group. Proceeds to force out current employees and replace them with his associates. Then abuse the review process to boost one of those to a higher position in the company. Playing the game right they all move up in 18 to 24 months. All fun and good except the groups they pass through are trashed in the process.

But the thing I don't understand is why Intel gave up on XScale, their ARM-compatible effort This isn't the first time Intel sold off a slice of itself. I remember AMD doing this many, many times. For over 40 years, the form factor didn't matter. All that mattered was having the fastest and most performant CPU in the entire world. There was a real demand to crunch numbers and move bits since the 70's, so that's the world Intel, AMD, and everyone else knew.

To sell off a division that sold low-power chips for tiny margins in an industry that never, ever, ever explodes (until it did) was just plain normal. He would have gotten pressure and funny looks from the industry and shareholders if he didn't sell it off.

Well by the same token AMD is entering the ARM chip business.

7 years ago they sold their Imageon GPU to Qualcomm, which is now going gangbusters inside every Snapdragon-based phone.

Adreno is an anagram of Radeon.

Atom is doing 'great'. Intel spend over $7 billion(1) subsidizing Atom in hopes chinese vendors would use it, they rarely did. Finally in 2015, partially driven by fear of prosecution for price dumping against native Chinese silicon manufacturers, they entered into a 'cooperation'(2) with gov run semiconductor giants, and by cooperation I mean Intel gave them 1.5B, free Atom license and committed to spend $5.5B on flash fabs in the mainland. Intel also also promised to shareholders they would stop subsidizing Atoms ... which didnt last that long (3).

1 http://appleinsider.com/articles/15/01/07/after-intel-spent-...

2 http://www.cnet.com/news/intel-doubles-down-on-mobile-with-1...

3 http://www.digitaltrends.com/computing/intel-to-chinese-elec...

Latest Atoms are almost usable, and you can get passable quad core 4GB ram/64GB flash tablet for $160 (Chuwi Hi10).

I believe the easy answer is that they thought they could adapt x86 to all markets, and that became the strategy.

More complicated is that Intel came into the mid-90s with several "next generation" architectures, some that weren't solely developed at Intel. XScale came from Digital and had IP from ARM. i960 was a joint venture with Siemens. And they still had the i860 which by this point it was clear was never going to meet expectations in the market. So when the x86 folks said "Intel should put all it's wood behind our arrow", politically they were fighting groups that were weakened by NIH or who'd already for all practical purposes failed. Probably didn't hurt that Pentium was doing quite good and P6 was looking good on the horizon.