Slashdot Mirror

← Back to Stories (view on slashdot.org)

Intel Launches Onslaught of Skylake CPUs For Laptops, Hybrids and Compute Stick

Posted by Soulskill on from the chips-and-dips dept.

MojoKid writes: Intel is following up on its Skylake launch bonanza by opening the floodgates on at least two dozen SKUs mostly covering the mobile sector. The company is divvying up the range into four distinct series. There's the Y-Series, which is dedicated to 2-in-1 convertibles, tablets, and Intel's new Compute Stick venture. Then there's the U-Series, which is aimed at thin and light notebooks and "portable" all-in-one machines. The H-Series is built for gaming notebooks and mobile workstations, while the S-Series is designated for desktops, all-in-one machines, and mini PCs. Also, the Y-Series that was previously known as simply the Core M, (the chip found in products like the 12-inch Apple MacBook and Asus Transformer Book Chi T300) is now expanding into a whole family of processors. There will be Core m3, Core m5, and Core m7 processors, similar to Intel's Core i3, Core i5, and Core i7 CPU models in other desktop and notebook chips.

54 comments

  1. Was this written by Intel? by Anonymous Coward · · Score: 0, Insightful

    Onslaught? Bonanza? Floodgates? Holy marketing Batman....

  2. Impressive by Anonymous Coward · · Score: 0

    They've made their already confusing product line-up even more confusing!

    1. Re:Impressive by Anonymous Coward · · Score: 1

      You are correct. We sell 68 different current Intel CPU models. That doesn't include all of the nonsocketed models that aren't end user installable like for mobile or embedded systems. There's probably over 150 different current Intel CPU models. Intel does seem to be trying to confuse the market on purpose.

    2. Re:Impressive by davester666 · · Score: 1

      Hello, market segmentation. Microsoft wishes they could do this.

      --
      Sleep your way to a whiter smile...date a dentist!
    3. Re:Impressive by PingSpike · · Score: 2

      Well, when most users can't tell the difference between any of your products (because there barely is any) and they aren't even worth the hassle of a motherboard upgrade from previous products (much less the cost of the hardware) creating new large confusing model numbers is one way to give the illusion of a purpose I suppose.

  3. 5K resolution by andymadigan · · Score: 2

    The only thing I want to know is:

    Will the 28W parts be able to drive a 5K display when used with Alpine Ridge (Thunderbolt 3)?

    That is: would a 13 inch Macbook Pro with Skylake be able to drive a Retina Thunderbolt Display?

    Supposedly Thunderbolt 3 does support 5K resolution, and the Intel Iris 550 SKU will have 64MB of eDRAM.

    I suppose we won't really know until next year.

    --
    The right to protest the State is more sacred than the State.
    1. Re:5K resolution by rsmith-mac · · Score: 3, Informative

      Will the 28W parts be able to drive a 5K display when used with Alpine Ridge (Thunderbolt 3)?

      Yes and no. Yes, they can. No, not in the way you want them to.

      Alpine Ridge only supports DisplayPort 1.2, which does not have enough bandwidth to drive 5K (you need DP 1.3). So instead Intel has it carry 2 complete connections (8 lanes).

      On paper that's enough bandwidth, but now you have to build a 5K display that uses multi-stream tiling to bond 2 interfaces. MST is kind of an ugly hack, and while Apple uses it on the 5K iMac since it's a closed system, it would be a bigger can of worms to use it on an external display given their demand for perfection.

    2. Re:5K resolution by Anonymous Coward · · Score: 0

      Doesn't the Dell 5K display require input from two DisplayPort streams? Now suppose that Dell revised their display to add a USB-C + Thunderbolt 3 input (from which the display internally extracted two DisplayPort 1.2 streams).

      Even if it was "kind of an ugly hack", wouldn't it appear like a tidy one-cable hookup to an end user?

    3. Re:5K resolution by rsmith-mac · · Score: 1

      Even if it was "kind of an ugly hack", wouldn't it appear like a tidy one-cable hookup to an end user?

      Yes. The issue isn't so much how it appears to the user as it is how it appears to the OS. MST displays have an unfortunate habit of having a tile drop out now and then, if only for a second. The iMac gets around this by being a closed system, but Apple would have to address this head-on with a 5K Thunderbolt 3 display. It's one thing for 3rd party monitors to do this, but it's another thing for 1st party monitors to do it.

  4. Re:Sounds like ... by Anonymous Coward · · Score: 0

    Well, they're a pretty small company, so I suppose that they have limited... They're freaking Intel..!!! Is there any large company that doesn't do this?

  5. Moore's new law by Waffle+Iron · · Score: 4, Insightful

    The number of distinct microprocessor SKUs on the market doubles every 18 months.

  6. It's no ARMv8 by Gravis+Zero · · Score: 1, Interesting

    Seriously, this inability to let x86 go is just getting sad. If you want something that is power efficient, you go with ARM chips. I wouldn't be surprised if Apple announced it was switching it's laptop/desktop machines over to their own ARMv8 chips because in addition to power savings, it wouldn't cost nearly as much as the chips from Intel.

    --
    Anons need not reply. Questions end with a question mark.
    1. Re:It's no ARMv8 by viperidaenz · · Score: 1

      Haven't Xeon CPU's had better performance per watt than ARM since like, forever?

    2. Re:It's no ARMv8 by Anonymous Coward · · Score: 0

      In case you haven't noticed, ARM performance sucks cock.

    3. Re:It's no ARMv8 by Anonymous Coward · · Score: 0

      Providing you count pre 2012 as forever.

      Intel loses on every battle field apart from single core performance.

    4. Re: It's no ARMv8 by Anonymous Coward · · Score: 0

      Aka the most important metric. ARM is shit an will never scale up the performance curve beyond tablets.

    5. Re:It's no ARMv8 by gl4ss · · Score: 3, Informative

      so what you're saying is that intel loses in every other metric except being faster? ok I might be able to see why people are sticking with x86 besides program compatibility.

      besides, I certainly hope that you're not trying to suggest that octacore arm is faster than an intel octacore when all cores are running calculation...

      arm loses on everything except watts, for certain things, and price. but even a chip for a cheap laptop would beat an arm chip priced for a laptop.

      I got a quad core 2.7ghz phone(and it's not a cheap phone! I could buy a decent laptop for the price) and my aging intel 2.5ghz laptop still beats it night and day, on a common web bench the slower by mhz intel being 2-8 times better than the arm chip..

      --
      world was created 5 seconds before this post as it is.
    6. Re:It's no ARMv8 by TheRaven64 · · Score: 1
      It depends hugely on the workload and it also depends a lot on the core. The ARMv8 ecosystem is quite diverse. For example, you have some players like nVidia's Project Denver, which fuses some of their GPU ideas with designs inherited from Transmeta. The Denver core is VLIW, but with staggered pipelines, so that results from one instruction in a VLIW bundle can be fed into the next (without needing rename registers, which are one of the biggest power sinks on a modern OoO CPU). When you start a program running, there's a simple decoder that turns ARM instructions into fairly inefficient VLIW instructions, but after a little while hot loops are optimised by a JIT and get a lot faster.

      At the other end of the design spectrum, Cavium's Thunder X has 48 ARMv8 cores (not hyperthreads) per die, and supports dual-socket configurations for up to 96 processors per board. Individually the cores are weaker than a Xeon, but on some workloads (network routing, some database serving), they're pretty impressive in aggregate. That many physical cores also makes it easier to load balance VMs in a hosted environment. This is especially good for the kind of workload where most clients are idle for a lot of the time, but when they're busy they're very busy.

      --
      I am TheRaven on Soylent News
    7. Re:It's no ARMv8 by arglebargle_xiv · · Score: 5, Interesting

      arm loses on everything except watts, for certain things, and price. but even a chip for a cheap laptop would beat an arm chip priced for a laptop.

      That was certainly something that came as a surprise to us about... oh, I think it was two years ago, just how underpowered ARM chips are compared to what look like similarly-specced x86 ones. We were looking at moving to at least some ARM-based server stuff for power and cost reasons, but quickly found that they don't come close to the performance of x86 hardware (that was after spending forever on tuning and optimisation, we assumed we'd set things up wrong but it turned out that they just don't make for good server devices).

      I'm not trying to bash ARM here, just pointing out that if I want to build a versatile tablet or embedded device I wouldn't think of anything other than ARM, but for a server I wouldn't think of anything other than x86 (or equivalent, Sparc, Power, whatever). They're just designed for totally different market segments.

    8. Re:It's no ARMv8 by Linuksistas · · Score: 2

      What is getting sad is this pointless x86 vs ARM discussion. It's like to discuss bike vs car. Both ARM and x86 have they own technological/economical strengths and weakness'es. We can't state some sort of absolute 'winner' here. ARM as architecture always had and probably always will have better power efficiency than x86. It's just inherited from primary ARM design points to be low-power embedded system processor first, no matter whether it'a LTE router or a smartphone. And it does it's job quite well. We should also mention the fact than most ARM cores comes as SoC's, so nearly every required peripheral is implemented in a single chip, while x86 more or less are still microprocessors only. Another important aspect is what partical ARM core do you have in mind? There are ARM reference cores (Cortex family), there are custom implemented ARM cores (like Apple, Qualcomm chips have), there (probably) will be some kind of custom/high-performance ARM implementations like ones coming (hopefully) in AMD K12. Yes, as for raw performance ARM may not be such microarchitectures as Skylake, but can easily outperform atom-based x86's. I personally believe that ARM (and ARM based SoC's) have enough performance to be employed in smartphone/tablet/ultrabook segments. However, it may not be ideal choice for full-blown desktop/workstation but it was not designed to be!

    9. Re:It's no ARMv8 by Anonymous Coward · · Score: 2, Interesting

      Thats always a funny one.

      Specially that it doesnt need any fancy argument. Just go to a phone shop and run the fastest, greatest phone or phablet with their UHD screen and octo core armv8 cpu, as well as their extremely well optimized version of linux.

      Now browse the web a bit.

      Now go get a cheap, current i5 laptop with an UHD screen. I'll just mention that this is magnitudes cheaper to get than the phone above.

      Now browse the web a bit.

      Ooh. Looky how much faster the laptop is. Now compare the size of the screen it has to power (biggest power draw). Now compare the power consumption at idle and running.

      My 2y old haswell laptop runs idle with screen on medium brightness at 3.1w. My phone which is a z3c with a 720p 4.6in screen runs right this moment at 3.3w with screen on, low brightness. armv7 that's supposedly using less power than the qcom armv8 incarnations.

      Here's where the phone wins: its using a few hundred of milliwatts when idle screen off. the laptop's equivalent is sleep-mode which no one sees as the same as phone screen off as desktop apps arent able to cope with being disconnected, and it takes a little longer to wake up the pc vs the phone.

      As far as load, the laptop goes around 10 watts in browsing vs 5w for the phone, and the laptop can peak at 20w if needed (its many times faster than at this power output).

      Now again why would I want a laptop with a slow-ass armv8 that consumes as much or more and isnt compatible with anything my pc runs?

      x86 not the x86 from 10 years ago. it works very similarly to arm (or in fact, arm works very similarly) except with more compatibility layers.
      Laptop

    10. Re:It's no ARMv8 by Anonymous Coward · · Score: 3, Interesting

      'xactly there's a reasons why OEMs don't go and make ARM laptops:

      - their power/speed ratio isn't what adverts claim compared to intel, nothing beat intel on on speed right now. nothing.
      - apps that run on x86 dont run on armv8, you need to be able to recompile, reoptimize, etc. which is often not possible.
      - non-xeon intel is iactually toe to toe with arm power-consumption wise.

      I think most people figure that if their phone lasts 1-2 day idle, arm must be the reason for it. its not though. its the OS, the screen, the ram, etc. but mainly, the OS. both arm and intel CPUs are very integrated nowadays and can fully turn off (ie use 0 power) for various components, very fast.
      On a phone, when screen is off, almost everything is powered off except for the RAM. apps data access is queued. apps wake lock (opportunity to wake up he cpu to run stuff) are queued. This is considered "idle".

      On a laptop, when considered idle, screen is on and everything is running. No wake lock for apps, its always on. No queueing.

      Yet under these conditions intel cpus consume a tiny fraction more than arms. Doesn't take a genius to figure out that under the same conditions the arm cpu will use as much or more power, or be slower. Worse! armv8 CPUs ain't even all that cheap.

      Thus as it stands, arm makes no sense for the workstation, laptop, server racks market.

    11. Re:It's no ARMv8 by Anonymous Coward · · Score: 0

      no. the sad thing is that arm is still trying to be what it's not.

    12. Re:It's no ARMv8 by Anonymous Coward · · Score: 0

      It's also completely ridden with both Intel and NSA spyware.
      ARM doesn't have any of that.
      Intel is not to be trusted anymore.

    13. Re:It's no ARMv8 by aliquis · · Score: 1

      Haven't Xeon CPU's had better performance per watt than ARM since like, forever?

      But.. But... the i3.. i286.. 8088 was bad!

      https://www.youtube.com/watch?v=yHXx3orN35Y

      .. forget about it.

    14. Re:It's no ARMv8 by Anonymous Coward · · Score: 0

      I think you're maybe being a little harsh on the price/perf comparison. Most phones have incredible displays. My first gen moto g looks killer. Go out and buy a $200 laptop and see how that turns out. Literally: Asus Chromebook (C300MA-DB01), 16GB eMMC, $219. Nope. Seriously, a reviewer on newegg says: "The screen has a narrow up and down field of view." I'll take the Moto G. Not everyone buys high end phones and high end laptops. At the low end especially, non-IPS laptops stink once compared to a phone, which almost always has a nice display.

    15. Re:It's no ARMv8 by Anonymous Coward · · Score: 0

      Sorry but your argument was lost somewhere in the early 2000s.

      Most other ISAs died because they were, frankly, inferior. The promises of RISC and VLIW turned out to be hollow because, as complexity increases, those "elegant" and "simple" ISAs fall in to a programming complexity trap. Turns out it's better to craft a complex, smart piece of silicon and implement the instruction there instead of make a smaller, faster chip and implement in software.

      Why? Because most code is crap. And no magic compiler will turn crap code in to good cpu instructions. You also need /more/ small instructions to do the same task and, if you have not noticed, we've reached a practical clock celing. In essence the complex instruction sets win because of bandwidth.

      Intel is the best chip maker in the world by a silly margin. They're a full 2 generations ahead of everyone else. They really do know what they're doing.

      And intel could not even slay x86. They tried. They made Itanium and it sucked so hard they had to license 64 bit instructions from AMD.

      ARM is great, but only at a few things. Single thread performance is bad and it turns out nobody really wants a dense brick of lots of slow single thread cores. Intel gave up on the lots-of-atoms-in-a-box idea and is now pushing the Xeon-D, which will sweep the dense server market. The Xeon-D is lots of low power but fast cores with real server features and real server throughput. It's also got a real head-turning instructions-per-watt that arm can't even touch.

    16. Re:It's no ARMv8 by nateman1352 · · Score: 1

      Both the A8X and the Broadwell Core M have a TDP of ~4.5W, so they give us a good comparison between the latest and greatest ARM vs. x86 CPUs:

      http://browser.primatelabs.com/geekbench3/compare/2959187?baseline=3338936

      Lets compare against the nVidia Tegra K1 as well, which has a TDP of 5W vs. the Core M's 4.5W:

      http://browser.primatelabs.com/geekbench3/compare/2959187?baseline=3347052

      As you can see, Intel is actually competing well against the best ARM can offer in their own backyard. The A8X does ~5% better in multi-threaded workloads, it has 3 cores vs. Core M's 2 cores. Single threaded, the A8X is ~26% slower than Core M. Despite having 4 cores, the Tegra K1 has ~16% worse multi-threaded performance than the Core M with 2 cores and ~53% slower single threaded performance. When you consider most GUI/web applications are single threaded (which is mostly what people use tablets for) Broadwell Core M is the best tablet chip on the market right now. Its only going to get better with Skylake.

      At the same time, ARM hasn't been able to really touch Intel's home turf in the high performance market.

      On the topic on instruction sets, honestly the most important difference between x86 and ARM is having an x86 design gives you a distinct advantage in the market for computers that run Windows. Given that there is no disadvantage for x86 in any other market segment (Android, Chrome, Mac, etc.) why would Intel switch to ARM when there is only upside to x86 and no downside?

    17. Re:It's no ARMv8 by Anonymous Coward · · Score: 0

      A 1 watt cpu by intel is worth it because backward compatibility...

  7. Back door? by Required+Snark · · Score: 1
    If you read the article, it says that Intel is becoming less and less specific about what is in their chips. For example, they are not providing a gate count for any of the Skylake chips. Additionally, the Power Control unit has a full CPU in it: "The PCU is essentially a microcontroller (we’ve seen references to a full Intel architecture (IA) core in there)".

    Given the above, there is an incredible opening for some TLA agency (N?A) to put their very own software/hardware back door in at the silicon level. How would anyone ever know? Given the fact that they have already tapped a large percentage of all the phones in the world, the logical next step would be to place bugging capabilities in every Intel CPU on the planet.

    Paranoid much?

    --
    Why is Snark Required?
    1. Re:Back door? by Anonymous Coward · · Score: 0

      Don't worry. It has been there for a while already

    2. Re:Back door? by kromozone · · Score: 1

      Other major chip manufacturers will conduct their own reverse engineering or hire speciality chip patent infringement firms to map and analyze new chips from competitors. So a company like Samsung will do a fairly thorough teardown of new Intel chips and have a fairly detailed map. Not that a suitably clever backdoor couldn't be hidden from this type of scrutiny, it's not exactly the type of thing they're looking for. The always-on Cortana DSP is worrying and they've admitted to that, that could certainly serve as a backdoor with some degree of plausible deniability. The data collection centers the NSA has now are probably sufficient for the type mass data collection they want and they don't have the risk of potentially sabotaging one of the biggest US corporations. There is some level of peer review in chip manufacture and I doubt Intel would want to sabotage themselves in that way, or that the US government could legally obligate them to sabotage their chips in that way.

    3. Re:Back door? by Anonymous Coward · · Score: 0

      If the NSA wants to spy on me, all they'll hear is porn audio and fapping.

  8. x86 isn't the performance bottleneck it once was by Theovon · · Score: 2

    x86 is no longer a microarchitecture. It's just an ISA. It's a total abstraction, and in mid-range to high-end processors, its translation overhead (logic and latency) is minimal. Only in the lowest-end devices (Atom) is it any kind of burden, and ARM dominates in that space.

    Yes, CISC is computersciencely evil, not orthogonal, crufty, and whatever else you want to call it. But these days, x86 is just an intermediate language between the compiler and the REAL execution engine.

  9. When's the core count ever going to change? by niceworkthere · · Score: 1

    Any chance we get to find out when Intel considers an increase in cores for these product segments?

    Because rather than a rather minuscule performance increase (compared to the hardware from 1-2 gens prior), that could actually make for a worthwhile reason to buy.

    1. Re:When's the core count ever going to change? by Required+Snark · · Score: 1
      Intel Phi: Knight's Landing.

      72 cores, 3 teraflops. Good enough?

      --
      Why is Snark Required?
  10. That's nice and all... by Anonymous Coward · · Score: 0

    But I still can't even find a desktop Skylake-S i7-6700K to save my life.

  11. Re:Sounds like ... by pushing-robot · · Score: 1

    Apple?

    --
    How can I believe you when you tell me what I don't want to hear?
  12. If it's anything like their 6700k "launch"... by cplusplus · · Score: 1

    ... you won't see it for months. The 6700k was a paper launch- lots of marketing with no product. The few you can get are marked up by 50% or more.

    --
    "False hope is why we'll never run out of natural resources!" - Lewis Black
  13. Re:x86 isn't the performance bottleneck it once wa by rahvin112 · · Score: 1

    You are wrong about the ATOMs. You don't understand that what's been wrong with Atom is intel's making, not some limitation of the architecture. Intel has deliberately handicapped Atom to avoid atom taking market-share from high priced parts. This lack of performance from Atom was completely deliberate, they wanted to make it painful to use it for anything so people would opt for the better and higher priced desktop parts.

    You want evidence? Intel has recently produced "server" atom's that are what Atom could be. They are the Silvermont and Avoton parts and they are good enough to run servers, 8 cores and a peak of 20 watts (at full load), where the idling power use is about 4 watts. Intel can produce power competitive parts that blow ARM out of the water, they don't because they don't want to erode the prices of their higher end parts. The Avoton I have can run circles around i7's that are just a year or two old and it does so at like 1/10th the power use.

  14. Anonymous Coward doesn't know how to use apostroph by Anonymous Coward · · Score: 0

    Film at 11

  15. Re:x86 isn't the performance bottleneck it once wa by U2xhc2hkb3QgU3Vja3M · · Score: 1

    Part number of your Avoton?

  16. Re:Sounds like ... by Molt · · Score: 1

    Apple have a TV-thing series, a watch series, a phone series, small tablets, larger tablets, small laptops, the large laptops, the desktop all-in-ones, and high-end desktop workstations. They are even apparently working on a car.

    This is exactly the same as Intel, different products for different spaces. Intel have not yet made a CPU which looks like a bin, though.

    --
    404 Not Found: No such file or resource as '.sig'
  17. Boycott Israel by Anonymous Coward · · Score: 0

    The new Intel chip was designed in Intel Israel.

  18. Re:x86 isn't the performance bottleneck it once wa by Anonymous Coward · · Score: 0

    Based on the TDP provided, I would guess C2750.

  19. Re:x86 isn't the performance bottleneck it once wa by rahvin112 · · Score: 1

    2758. Same TDP slightly better performance.

    http://www.supermicro.com/prod...

  20. Re:x86 isn't the performance bottleneck it once wa by Blaskowicz · · Score: 1

    That may be because a server doesn't need a shit ton of CPU, and hardware AES helps.
    The other aspect is your server Atom CPU has quite unrestricted power use, comparatively.
    See, this quad core Atom (named Celeron) has a GPU, is constrained to about a third the power use and thus has to underclock itself.
    http://ark.intel.com/fr/produc...

    Constraining any CPU to well below 10 watts will make it suck (unless you're satisfied with the CPU power. Netbooks are quite good if they do what you want of them)