Intel maintains steady growth across the board in Q1 2018

Hot on the heels of AMD's healthy earnings report yesterday, Intel posted strong Q1 2018 results of its own this afternoon. The blue team took in $16.1 billion in revenue, up 9% year-on-year. The company turned a $4.5 billion operating profit on that revenue, up 23% on the year. Net income was $4.5 billion, up 50% on the year, and earnings per share of 93 cents was up 53% on the year. Gross margin fell 1.3 percentage points on the year, to 60.6%.

  Q1 2018 Q1 2017 Change
Revenue $16.1 billion $14.8 billion up 9%
Operating income $4.5 billion $3.6 billion up 23%
Net income $4.5 billion $3.0 billion up 50%
Gross margin 60.6% 61.9% down 1.3 pp
Earnings per share $0.93 $0.61 up 53%

Intel's PC-focused Client Computing Group brought in $8.2 billion in revenue, up 3% from this time last year. Operating income fell 8% to $2.8 billion. PC platform volumes (a measurement of processor and chipset sales) were flat on the year, though desktop platform average selling prices rose 7% and notebook average selling prices rose 1%. Interestingly, Intel attributed the fall in CCG operating income in part to its 10-nm transition efforts.

The blue team trumpeted its data-centric business results in its earnings release, and it's easy to see why. Revenue from the Data Center Group rose 24% on the year, to $5.2 billion, and the group's operating income rose 75% to $2.6 billion. Data center platform revenues rose 24% on the year, and unit volumes grew 16%. Average selling prices of data center platforms rose 7%. Among the distinct market segments Intel cites, cloud service provider revenues grew 45% and communication service provider revenue grew 33%. Government and other enterprise revenue grew 3%.

The Internet of Things group posted $840 million in revenue and $227 million in operating income, a 17% increase in year-on-year revenues. The Mobileye business unit contributed another $151 million in revenue and $10 million in operating income to the IOT division. Intel cites "retail and video" as the drivers of this growth. The Non-Volatile Solutions Group enjoyed revenue of $1.04 billion, though the division still posted an $81 million loss (compared to a $129 million loss a year ago). The Programmable Solutions Group posted $498 million in revenue, up 17% on the year, and made $97 million in operating profit.

For its second quarter of 2018, Intel expects $16.3 billion in revenue on a GAAP basis, plus or minus $500 million, which would represent a 10% increase from the year-ago quarter. The company expects operating margin of about 28% and earnings per share of about $0.85.

Comments closed
    • strangerguy
    • 2 years ago

    Intel’s main problem is their overemphasis on their old school brute-forcing node advancements to maintain their process lead with increasingly questionable real benefits rather than using the same resources to develop better chip designs. As early as 2012 Apple was already starting to prove their old approach isn’t going to work much longer.

    • Srsly_Bro
    • 2 years ago

    let me give a shoutout to my boi @chuckula to put his Intel masters on notice.

    [url<]https://www.tomshardware.com/news/intel-cpu-10nm-earnings-amd,36967.html[/url<] Intel got 11 problems and all but one in nanometers.

      • blastdoor
      • 2 years ago

      Good article. This quote in particular jumped out at me:

      “Intel’s relatively flat R&D spend (+3%) is certainly not encouraging, given the current climate.”

      Sounds like they are losing their lead primarily because they aren’t willing to pay the price of keeping it.

      • chuckula
      • 2 years ago

      I was way ahead of you yesterday in CONFIRMING that Intel will go bankrupt before they ever ship a 10nm part.

      THG should pay me a finder’s fee.

      • blastdoor
      • 2 years ago

      An older article, but still relevant:

      [url<]https://www.semiwiki.com/forum/content/6713-14nm-16nm-10nm-7nm-what-we-know-now.html[/url<] At the time this article was written, it was believed that Intel's 10nm would be competing with TSMC 10nm, and of course Intel would win that competition by a landslide. TSMC's 7nm would put them ahead of Intel's 10nm, but it was believed that TSMC's 7nm would come out a fair bit later than Intel's 10nm. So in terms of timing estimates, this article is obviously out of date. The density estimates, though, seem relevant.

        • Srsly_Bro
        • 2 years ago

        Any comparison to glofo’s 7nm? AMD released a teaser saying 7nm instinct is in its labs.

        Nevermind..

        It looks like it is tsmc 7nm.

          • blastdoor
          • 2 years ago

          Here ya go:

          [url<]https://www.semiwiki.com/forum/content/7191-iedm-2017-intel-versus-globalfoundries-leading-edge.html[/url<] Edit: Comparing the "Intel metric" for density across these articles, what we get is: Intel 10nm == 103 TSMN 7nm == 116.7 GloFo 7nm == 90.5 (higher numbers are better) edit 2: for comparison sake: Intel 14nm == 43.5 TSMC 16nm == 28.2 TSMC 10nm == 60.3 Holding marketing labels fixed, Intel wins. But if we look at contemporaneously shipping products Intel is already losing because TSMC 10nm is superior to Intel 14nm. By these numbers, there is no evidence that TSMC will trail Intel again. Of course, this is just density, and there are other characteristics that matter. The proof will be in the pudding -- what products can be made on these processes? But there's no way this is good news for Intel. I'm surprised their stock price hasn't fallen on the news of the 10nm delay.

            • chuckula
            • 2 years ago

            I don’t see why you are so hung up about these processes when the Apple 2020 Miracle ARM chip is — shall we say — [i<]confirmed[/i<] to be on a 5 nm process that annihilates these obsolete "10nm" and "7nm" processes anyway. I've already placed my preorder for my iPad that is guaranteed to destroy Tiger Lake X.

            • Srsly_Bro
            • 2 years ago

            Maybe that’s why Intel 10nm is delayed.. It’s “dead in the Fab” and Apple is going to buy Intel so it (Apple) can make it’s own chips.

            • chuckula
            • 2 years ago

            Apple is about to buy Intel…. wait for it….. [I’ll post the rest in 3 months]

            • blastdoor
            • 2 years ago

            Why would Apple buy a laggard like Intel?

            BTW, Intel needs to get their yields up on LTE modems for Apple on their 14 nm process. TSMC appears to have no problems fabbing for Qualcomm on their 10nm process. If Intel’s fabs can’t hack it, then they are going to have to go back to using TSMC again, like Infineon used to.

            • chuckula
            • 2 years ago

            Good good.
            Let the [s<]hate[/s<] [u<]RDF[/u<] flow through you. Did I mention that the Apple 2020 Miracle ARM chip has the modem built right in? Soon Apple will have no need of puny "suppliers" for their products.

            • blastdoor
            • 2 years ago

            [url<]https://www.fastcompany.com/40564327/intel-to-supply-70-of-iphone-modems-this-year-100-in-2019[/url<] [quote<]As of now, our source says, the modem-chip yield rates are not what Intel expected. Only just more than half of the chips being produced are keepers. The Intel engineers are confident they can work the bugs out and push up the yield rates before production ramps into high gear in June and July. Until this year, TSMC had fabricated Intel’s modem chips (using its 28 nanometer process). Intel originally acquired its modem-chip business when it bought Infineon in 2010. TSMC had been Infineon’s fabricator, and Intel kept the arrangement going after it acquired the modem. [/quote<]

            • chuckula
            • 2 years ago

            This is great news considering demand for the iPhone X has cratered and Apple is about to fire-sale them.

            We can clearly blame the failure of the $1000+ phone on Intel and redouble our efforts for the 2020 Miracle ARM chip!

            • blastdoor
            • 2 years ago

            Intel’s low yields are good news because iPhone X sales are low?

            I think you are desperate to change the subject.

            But that’s fine with me.

            Regarding iPhone X — I personally think Apple made a mistake in the feature mix. I’m perfectly happy with the screen on my iPhone 7+. Paying a premium for an OLED screen does not interest me. But there is still a lot of room to improve photography on smartphones and I would be willing to pay more for a better camera. If I had been on the market for a new phone last year, I would have bought the 8+ rather than the X. Or I might have said screw it, I’ll hold on for another year and wait for them to straighten out their product lineup.

            • chuckula
            • 2 years ago

            [quote<]Intel's low yields are good news because iPhone X sales are low? [/quote<] Tsk, tsk, you forgot the RDF: [ENGAGE RDF]: iPhone X sales are low [b<]because of Intel's low yields![/b<] Therefore, this is good news for Apple kicking Intel out of all its products! You need to start holding it right.

            • Srsly_Bro
            • 2 years ago

            Thanks for the info, bro. I wasn’t aware of all that.

    • moose17145
    • 2 years ago

    Another unsurprising quarter for Intel. Nothing to see here… move along.

    I say that in all seriousness. Despite owning a decent amount of Intel stock, I truly believe it would have been nice to see AMD put a legitimate dent in Intel now that they are competitive again. But (as long as AMD keeps their current momentum going), I suspect it will take a few more quarters for the Zen architecture to start putting actual noteworthy dents into the 800 pound gorilla that is Intel.

    To be completely honest… I would LOVE to see a world where AMD and Intel are standing on roughly equal footing where the CPU market is divided close to 50/50. Heck maybe 45/45 with Via bringing up the remaining 10% of the x86-64 market (lots of people forget they also have a x86 license). Competition is good for the consumer and helps drive innovation! But, innovation can only be driven so long as their is real competition WHILE providing competitors with the profits needed to research and develop those innovations.

      • ronch
      • 2 years ago

      I reckon VIA’s license will expire soon (or maybe it already has) and there’s no word that they’ll extend it.

        • moose17145
        • 2 years ago

        I’m not so sure about that… The inquirer had an article from Jan this year stating Via is planning on bringing out some new x86-64 chips sometime soon(tm). Not sure how TR missed that… or maybe I was the one who missed it.

        [url<]https://www.theinquirer.net/inquirer/news/3023729/via-technologies-to-launch-x86-64-cpus-in-bid-to-take-on-amd-and-intel[/url<]

        • tipoo
        • 2 years ago

        I looked at that a while ago, hard to know what’s going on with them. Their licence has been about to expire but renewed a couple of times now, I recon Intel doesn’t remotely worry about them and is happy to extend the licence for a while as a hedge against non competition. But Ryzen seems to have caught them unaware, so maybe they won’t feel the need to keep renewing VIA’s licence.

    • blastdoor
    • 2 years ago

    If NASA were trying to ascertain the existence of AMD using these numbers, they might infer it from the wobble in the gross margin. Otherwise, it would be pretty hard to spot.

    • NTMBK
    • 2 years ago

    10nm high volume manufacturing is officially delayed into 2019.

      • blastdoor
      • 2 years ago

      And TSMC recently started high volume manufacturing of their 7nm node.

      2018 — the year Intel’s process advantage was lost.

        • Gadoran
        • 2 years ago

        Pretty certain they does not manufacture high power cpus neither laptops SKUs but only very low power SOCs for phones.
        Sure you understand the difference on a so compact process

          • blastdoor
          • 2 years ago

          Excuses, excuses.

          And not even accurate excuses.

          Core M is a 4.5 watt or lower part. Apple’s SOCs are definitely in the same range, higher for the iPad.

            • Gadoran
            • 2 years ago

            Ummmm a Phone SOC is below the 2W power budget.

            Do you mean iPad SOCs?? They are not at work for them.

            Anyway a september 10nm Core M is very likely but not necessary for Intel balance sheet.
            Why to ship 10nm laptops or desktop cpus if they can not run at 4.5/5Ghz turbo with high yields ? Better the old process with some refinements more like has done GloFo for AMD.

            A new process has a sense only if you can do economics on it with your SKU line.

            • blastdoor
            • 2 years ago

            2 watts? You’ve clearly never attempted to record 4k video on an iPhone 7. That sucker gets hot. There’s no way 2 watts is doing that.

            But yes, you’re right — the economics of a new process don’t work out for Intel, which is why they are holding on to 14 nm +++++.

            And why don’t the economics work out?

            Their total and complete failure (perhaps refusal is a better word, actually) to compete in the biggest market for integrated circuits in the history of the industry.

            Intel better just get down on their knees and pray that GloFo fails to bring out their 7nm next year. Otherwise, Intel is going to be stuck competing with AMD on a level process playing field. I don’t think that’s ever happened before. Good thing Intel hired Jim Keller — he’s a master at designing his way out of the constraints of an inferior manufacturing process.

            • Gadoran
            • 2 years ago

            You have not clue about manufacturing, this is pretty clear. Useless conversation on these basis.

            About phones, 2W is a golden standard to give a decent life to battery and to the screen, still under heavy loads the power goes up but the battery go down un few minutes without mercy. Your loved phone very hot on your hands. and your battery pack destroyed by a temperature over the absolute limit of 40°C, real killer situation for battery health.

            The reason of the failure of battery packs is that idiotic users don’t understand they have a phone in their hand, not a pc.

            • cegras
            • 2 years ago

            [quote<]The reason of the failure of battery packs is that idiotic users don't understand they have a phone in their hand, not a pc.[/quote<] Placing the onus on the user to preserve their phone is a terrible engineering decision. You have no idea what you're talking about.

            • the
            • 2 years ago

            Intel half heartedly tried to compete in mobile. They just couldn’t get past the realization they needed to lower their margins and make up the profit in higher sales volume. By the time Intel figured this out, they had already lost.

            Random related aside, I came across a box for Broxton software development platform. So Intel had at least sampled these 14 nm ultra mobile parts before [url=https://www.eetimes.com/document.asp?doc_id=1329580]formally cancelling them.[/quote]

          • Leader952
          • 2 years ago

          Pretty certain you are wrong.

          AMD has 7nm Radeon Instinct running in its lab
          [url<]https://www.fudzilla.com/news/ai/46157-amd-has-7nm-radeon-instinct-running-in-its-lab[/url<] TSMC Kicks Off Volume Production of 7nm Chips [url<]https://www.anandtech.com/show/12677/tsmc-kicks-off-volume-production-of-7nm-chips[/url<] [quote<]Unlike TSMC’s CLN10FF, which is used by a limited number of customers for a limited number of mobile SoCs, the CLN7FF is expected to be used to build CPUs, GPUs, FPGAs, neural network processors, cryptocurrency mining accelerators, mobile SoCs and so on. [/quote<]

            • Gadoran
            • 2 years ago

            Bet i am right?? 🙂

          • NoOne ButMe
          • 2 years ago

          Yeah, to bad thee are no high performance processes like, oh, ones used for GPUs. Or anything based on IBM’s technology with products already taped out…

          Oh, wait…….

            • blastdoor
            • 2 years ago

            In fairness, taped out is not the same thing as volume production.

            But he set the benchmark as laptop CPU power without realizing that a high-end smartphone SOC will hit power usage similar to Core M for brief (and sometimes not so brief) periods of time.

            As I said, try recording 4k video on an iPhone 7 and claim with a straight face that the SOC is only using 2 watts.

            When 14nm first rolled out, it was with the lowest power Core M processors. It took Intel a while to roll out the higher power stuff.

            Intel can’t even produce Core M in volume on 10nm. That’s how bad things are.

      • derFunkenstein
      • 2 years ago

      Yeah that’s a big yikes.

      [url<]https://twitter.com/markhachman/status/989598365003415552[/url<]

        • Pwnstar
        • 2 years ago

        Wow! Well, there you go.

      • the
      • 2 years ago

      This is really bad news.

      Cannon Lake is effetely still born. Technically it is ‘shipping’ but only volume enough seemingly that some executive could claim a bonus. No known products to date using these chips. Speaking of, Cannon Lake is being offered in a variant without IGP presumably due to yields. This is really bad as Intel has at least always offered GT1 level graphics so that a fully functional system could be built around the SoC.

      TSMC has caught up and it looks like Global Foudries will too. Intel used to have an 18 month lead in process technology but it has utterly disappeared.

    • Shobai
    • 2 years ago

    Frost! [Is that how this is done, SSK et al.?]

    Well done, Intel – celebrate the small successes.

      • chuckula
      • 2 years ago

      Intel hangs on by its fingernails for one more quarter.
      But what happens when the crypto bubble bursts and 90% of the sales go away?!?

        • moose17145
        • 2 years ago

        Nothing.

        Unlike NVidia, and especially AMD on the graphics front, Intel is not really as subject to the whims of the cryptomining mining market.

        Now if Intel had a graphics division (like an ACTUAL graphics division with discrete videocards)… In THAT hypothetical situation, Intel as a company would be fine, but the graphics division specifically would likely be quite subject to the whims of the cryptomining market. This is working under the assumption that Intel graphics cards would be as capable as NVidias and AMDs in terms of mining.

        But the CPU market itself I would argue is largely unaffected by the whole cryptomining thing. The graphics market is affected so bad only because each person doing mining is buying 6 or more cards at a time. But if you buy the right motherboard you can control 15 or 16 graphics cards with a single bottom of the barrel CPU. Hence why there has never been a CPU shortage throughout all of the cryptomining bubbles that seem to happen (roughly) annually.

        I would be willing to bet that even if we looked at AMD’s CPU division (specifically their CPU division), that even they were not too overly affected by the crypto craze and that their profits in that sector have been the result of some engineers that deserve some serious praise for bringing a legitimately competitive product to market on a shoestring R&D budget (relative to Intel). Again… their graphics division is an entirely different discussion…

          • K-L-Waster
          • 2 years ago

          Uh, I think your sarcasm detector needs recalibrating…

            • moose17145
            • 2 years ago

            Entirely possible. That thing doesn’t always function correctly in the mornings when I am checking this stuff.

            Edit because early morning grammatical errors… *goes and fetches more caffeine*

Pin It on Pinterest

Share This