Intel reports Q2 financials, qualifies Skylake chips for production

Intel has released its financial results for Q2 2015. The company reported operating income of $2.9 billion on revenue of $13.2 billion, down 25% and 5% respectively from a year ago. Gross margin fell two points year-over-year to 62.5%, while earnings per share remained flat at 55 cents.

Here's a tabular summary of the results:

  Q2 2015 Q2 2014 Change
Revenue $13.2 billion $13.8 billion down 5%
Operating income $2.9 billion $3.8 billion down 25%
Net income $2.7 billion $2.8 billion down 3%
Gross margin 62.5% 64.5% down 2 points

Performance results by group were mixed. The Client Computing Group took in $7.5 billion, down 14% from a year ago, and the company's software and services businesses saw revenue of $534 million, down 3% from this time last year. Other sectors had brighter results: the Data Center Group brought in $3.9 billion, up 10% from a year ago, while the company's Internet of Things business brought in $559 million, up four percent year-over-year.

CEO Brian Krzanich pointed to the results as evidence of the company's broadening focus, noting that 70% of operating profits came from growth in the data center, memory, and Internet of Things spheres. Those businesses helped to offset a soft PC market.

Looking forward, the company reports that it has qualified its sixth-generation "Skylake" Core processors for production during the second half of the year. Along with these chips, Intel thinks that the release of new PCs with Windows 10 will "bring excitement to client computing" later this year. For Q3 2015, the company forecasts revenue of $14.3 billion, plus or minus $500 million, and a gross margin of 63%, plus or minus "a couple of percentage points."

Comments closed
    • Ninjitsu
    • 4 years ago

    Worth noting that Client Computing revenue is +2% over Q1 2015, so it’s not all bad.

    • ronch
    • 4 years ago

    By the way, I don’t recall ever seeing an in-depth official article about the inner workings of Skylake. More pipelines? Bigger caches? Smarter schedulers? What?

    Intel has been unusually silent about its upcoming CPU.

      • chuckula
      • 4 years ago

      Nobody knows. About the only confirmed detail is that some versions (maybe not desktop parts though) come with AVX-512 support.

      The lack of info about Skylake is not encouraging since Intel typically has no problems playing up new features prior to launch.

      • gabest
      • 4 years ago

      Let me make a guess. It’s a sandy bridge, but with even more transistors wasted on a GPU that is good-enough to boot the bios, but not to play any games released in the last three years.

        • Chrispy_
        • 4 years ago

        I see the point you’re making but it’s possible to play recent games on even very low-power hardware.

        A great example is the Broadwell Y Core M processors with middle-of-the-road GT2 graphics, and at really quite low clocks to fit into the 8W or 13W thermal envelope.

        By far the most popular gaming category is MOBAs at the moment and games like DotA2 and HotS run very well on these integrated graphics, but even AAA games like BF4 run well enough – here’s a guy [url=https://www.youtube.com/watch?v=wbKvJHyqss0<]playing on a BF4 server using Core M's GT2 graphics[/url<] and getting 20-30fps even when the action intensifies. It's not dGPU performance, but it's playable. If you look at uploads from [url=https://www.youtube.com/channel/UCXIuoZK11s9jC_L27KgN5TQ<]SurfaceProGamer[/url<] You'll see that plenty of stuff runs even better than this - eg CoD:AW, and slightly older games like Fallout New Vegas are bumping up against the 60Hz Vsync limit. For those of use with dual-slot, several-hundred dollar graphics cards this is irrelevant, but three laptops are sold for every desktop, and even among desktop variants, things like NUCs and HTPCs exist where the integrated graphics genuinely matter. They'll be hooked up to TVs and in student dorms around the world.

    • ronch
    • 4 years ago

    From an earlier article here on TR:

    “AMD will announce its Q2 financial results next Thursday after the markets close, and it’ll hold a conference call that day to discuss the results at 2:30 PM PT.”

    • TopHatKiller
    • 4 years ago

    I totally agree. I’m excited about the increase of client market. Not to mention the IoT stuff from nowhere to nowhere. Let’s think, the Evil Empire has managed to hold it’s own [sort of] with no competition at all. Those Intel guys are so frigging great. Let’s all have a party for them!!!

      • NoOne ButMe
      • 4 years ago

      I like to read it as “we had well over a 70% of operating profit from data center and memory. We lost a lot of money paying people to take our IoT offerings”.

        • chuckula
        • 4 years ago

        IoT is a profit-making division.
        Don’t confuse smartphone chips with IoT devices, because they aren’t the same thing.

        When I was in grad school I went to a talk by Pat Gelsinger — who was still at Intel back then — who gave us this rather silly sounding and grandiose vision of distributed devices with pervasive wireless, sensors everywhere, etc. etc.

        It sounded awful silly… considering it was late 2003 and the main products that Intel had on the market were Pentium 4s.

        Interestingly enough, more than 10 years later, it doesn’t sound so silly. Almost like people at Intel think long term… gee I wonder why they manage to keep churning out those consistent profits?

          • TopHatKiller
          • 4 years ago

          Dearest Mr.Chuckerless, That’s lovely for you, honest, when I was at school [distant days – full of agony!] no-one of any note ever came to talk. However, it was not prescient of Mr.Gelsinger to chat about “IoT” or the comm-distributed / surveillance world. SF Writers have been discussing the whole concept for…20 years? 30? The “IoT” concept was not devised by Intel, and they weren’t planning for twelve years to implement their grand world-wide strategy. That’s just silly.

    • ronch
    • 4 years ago

    PCs and other computer form factors are increasingly becoming like refrigerators and ovens. These household appliances are usually in service for as long as they’re working. With PC performance generally increasing less and less with every new product release, people just don’t find enough reasons to upgrade. Same with smartphones and tablets.

    Big chip companies and everyone else associated with computers will inevitably need to cut production and personnel to adjust to these market trends.

      • odizzido
      • 4 years ago

      Or release a product worth upgrading to. They keep pushing more and more resources into their IGP which I just don’t care about. If they were to worry more about their x86 cores perhaps I wouldn’t be running a CPU from 2009.

        • NoOne ButMe
        • 4 years ago

        They don’t have much room to push IPC up on such a mature architecture. It’s very hard.

        • ronch
        • 4 years ago

        That’s more applicable to AMD than Intel. Improving GPU performance on a desktop is a simple matter of adding a discrete video card to the system in the case of APUs, but one would be stuck with whatever CPU cores AMD puts in their APUs. And yet they’re pretty much at a standstill in terms of CPU performance and keep giving us more GPU resources like we’re crying out for it.

    • HisDivineOrder
    • 4 years ago

    I don’t think Windows 10 is going to boost sales of PC’s much, if at all. Nothing about Windows 10 requires an upgrade for people with vaguely modern PC’s. And a free upgrade is going to be the big story with it, not the need for new hardware.

    Especially if a LOT of people already have Windows 7 or 8-class hardware.

      • omf
      • 4 years ago

      Yep, if anything the free upgrade is bad news for hardware makers, as many people took the opportunity to get a new machine rather than just buy the OS and deal with the upgrade process.

    • TwoEars
    • 4 years ago

    They still have a gross margin of 62.5%, down 2% from 64.5% last year, so they’re not exactly doing poorly.

    But they could be facing more competition in 2-3 years. Arm is still there and getting better, AMD has new leadership (and an engineer at that), IBM has been busy in the basement and could be licensing new interesting tech to TSMC. Both AMD and Samsung are likely interested in what IBM and TSMC are brewing and so forth.

    Could be interesting to say the least.

    • chuckula
    • 4 years ago

    For the usual crowd, while this quarter was certainly nothing to jump up & down about it also beat the apocalyptic predictions so the stock is doing pretty well: [url<]http://www.cnbc.com/2015/07/15/intel-q2-earnings-2015.html[/url<] Basically this quarter is Intel showing that while the PC market is certainly still important to Intel, they are also not defined by the PC market and can continue to pull in 2+ Billion in profits each quarter even when the PC market is in the dumps.

      • Meadows
      • 4 years ago

      For what it’s worth, both intel and AMD stock have been falling throughout July so far, although there’s a value difference of an order of magnitude.

      • Leader952
      • 4 years ago

      [quote<]Shares of Intel surged as high as 8 percent in extended hours. [/quote<] And at this time Intel is only up 1%. [url<]http://finance.yahoo.com/q?s=INTC&ql=0[/url<] Edit @ 12PM EDT: Now it is DOWN 6 cents (-0.2%) Edit #2 @ 1:43PM EDT: Now it is DOWN 50 cents (-1.67%) Edit #3 @ Market Close: Finished UP 21 cents at $29.90 (0.71%). Days trading range was 29.10 - 30.48.

    • chuckula
    • 4 years ago

    [quote<]CEO Brian Krzanich pointed to the results as evidence of the company's broadening focus, noting that 70% of operating profits came from growth in the data center, memory, and Internet of Things spheres. [/quote<] Basically: The PC can die but Intel won't.

      • ronch
      • 4 years ago

      If there’s one company that has demonstrated time and time again that they will never die, it’s AMD.

        • Deanjo
        • 4 years ago

        By selling off their assets. That well is pretty much dry.

    • jjj
    • 4 years ago

    In the call Intel confirmed a far bigger than expected delay for 10nm by saying Cannonlake will arrive in the second half of 2017! So TSMC and Samsung will likely be significantly ahead.
    While at it they also killed Moore’s Law by saying that it’s more like 2.5 years than 2 years now.

    Will edit later to comment on results but Cannonlake’s timing is big news, couldn’t wait.

      • chuckula
      • 4 years ago

      [quote<]So TSMC and Samsung will likely be significantly ahead.[/quote<] LMFAO. No, they really aren't. I don't care if Samsung literally launches what it calls a "10nm" SoC tomorrow, they still aren't ahead because what they call 10nm and what Intel calls 10nm are horses of very different colors.

        • jjj
        • 4 years ago

        Even if the size would be slightly bigger ( not that we actually have numbers yet), size is not everything and you know it but you can’t accept it Anyway go ahead and troll, i can just ignore your spam.
        If they use QWFET on 10nm ,it would be interesting but drop the fanboy BS and go for a reasoning that stands.
        It’s amazing how bad comments here got here lately. 80% are children with nothing relevant to say. It’s a wonder they can use a damn keyboard.
        Fed up with Intel fanboys, AMD fanboys, Nvidia fanboys , Apple fanboys , it’s a circus, you can’t have a freaking conversation anymore because this is their playground and they feel they can act as stupid as they are.
        Feel free to 403 m.

          • nanoflower
          • 4 years ago

          Okay, how about this for a reasoning that stands. TSMC has yet to deliver on time in recent years so there’s little reason to expect that they will get their version of 10NM out on schedule.

        • K-L-Waster
        • 4 years ago

        C’mon, Chuck, they really want to see Intel go out of business before AMD does – let’em have their fun.

          • chuckula
          • 4 years ago

          It could happen!

          You know how they had a Pentium Anniversary Edition last year?
          Well it’s been 10 years (10 YEARS MAN) since Prescott really went into full bloom in 2005.
          If Skylake is just a 10 year anniversary edition of the P4 then maybe Intel really will go under first!

        • Srsly_Bro
        • 4 years ago

        *horses of different sizes

          • ronch
          • 4 years ago

          Chuckie was obviously referring to My Little Pony.

            • chuckula
            • 4 years ago

            Apparently nobody has ever seen the Wizard of Oz?
            [url<]https://youtu.be/DJPCqUhSRbY?t=1m19s[/url<]

        • NoOne ButMe
        • 4 years ago

        Yes, previous to 14nm Samsung’s 10nm would have meant fully 10nm.

        Now it will mean that it isn’t really 10nm just like Intel’s 10nm isn’t really 10nm.

        yawn.

        Anyhow, Intel is still the most likely to launch a mass production for chips that aren’t ultra-high margin unless someone licenses something that IBM designs.

        • TheFinalNode
        • 4 years ago

        [url=https://www.semiwiki.com/forum/content/3884-who-will-lead-10nm.html<]Here's an analysis[/url<] about this worth checking out, which compares the density of processes from different foundries. For a long time, the foundries ramped up their new nodes a year or two after Intel but they were a lot denser than Intel's. That changed at 14nm, where Intel shrunk both the front and back-end-of-line, while the foundries implemented FinFET transistors to the previous 20nm back-end-of-line. It's not just about density though, but how the transistors perform. I think Intel will continue to have an advantage there because their processes are designed for their processors, while foundries have to cater to many customers. Plus, System-on-Chips are more challenging to manufacture, from what I've heard. Something about many different IPs being integrated onto one die. [url=https://www.semiwiki.com/forum/content/4801-who-needs-lead-14-10-7nm-nodes.html<]Here's another article[/url<] from a few days ago, discussing the timing of future processes.

      • brucethemoose
      • 4 years ago

      I don’t understand. We don’t even have 14nm stuff from TSMC/Samsung yet, but Intel is selling big, high-power 14nm parts right now… What makes you think they’ll hit 10nm before Intel?

        • Ushio01
        • 4 years ago

        If you want a Samsung 14nm chip go buy a Galaxy S6 smartphone.

        • NoOne ButMe
        • 4 years ago

        They’re spending more money on it.

      • Convert
      • 4 years ago

      I don’t really see them being all that comparable. I mean, sure, someone can claim 10nm first but unless the chips are more directly comparable than a smart phone SOC and a transistor behemoth I don’t really see the point.

      Unless you mean significantly further ahead of being behind than they were before? That I could see. If Intel is having a problem switching to 10nm I can’t see someone like TSMC nailing the execution, it would be a first if they did. But It’s definitely possible the gap is going to close a bit more this time around.

      I was actually thinking about what you are talking about earlier this week. Really we don’t have much more to go with current technologies. At some point all competitors are going to be on pretty equal footing in process tech. They will all have their little special twists but it’s going to be pretty interesting to see this become less of a differentiating factor. Maybe if AMD can hold on a bit longer they might get a second chance haha.

Pin It on Pinterest

Share This