Topic: Chips

Each post below is tagged with
  • Company/Division names
  • Topics
  • and
  • Narratives
  • as appropriate.
    Microsoft Announces Details Around Windows 10 on ARM (May 12, 2017)

    This content requires a subscription to Tech Narratives. Subscribe now by clicking on this link, or read more about subscriptions here.

    Nvidia Announces Strong Earnings for March 2017 Quarter (May 9, 2017)

    This content requires a subscription to Tech Narratives. Subscribe now by clicking on this link, or read more about subscriptions here.

    Intel Has a Serious Hijacking Flaw in its Chips, PC OEM Patches Coming Soon (May 8, 2017)

    This content requires a subscription to Tech Narratives. Subscribe now by clicking on this link, or read more about subscriptions here.

    Intel Reports Good Growth and Higher Margins, But Data Center Falls Short (Apr 28, 2017)

    Intel had a pretty good quarter for the most part, expanding margins as several of its divisions performed better off the back of rising prices and lower unit costs. The only division that didn’t grow was the security division (home of McAfee), which Intel has already announced it intends to spin off. Client Computing, home of Intel’s traditional PC business but also newer areas like tablets, phones, modems and so on, grew by 6%, though in large part because of rising prices (up 7% year on year) rather than volumes (down 4% year on year). The trend in Data Center, which makes up a little under a third of revenue, was similar, with volumes down 1% but prices up 6%, driving overall growth, though at a slower rate than in the past, and its margins have also fallen recently as it invests in new processes which should pay dividends over the longer term. Intel’s non-volatile memory group benefited from the release of Intel’s new Optane memory products, which have contributed to very rapid growth, but which has also plunged that division into the red because of heavy investment in the new processes and technology.  IoT is also growing at a decent clip, though it’s much smaller than the two big divisions. One other division is Intel’s Programmable Systems Group, which was formed when it acquired Altera, which made filed-programmable gate array (FPGA) technology. That’s growing at a decent clip and is also now profitable, which is good progress from when the acquisition closed. Overall, Intel’s two main businesses are performing well for it at the moment, though the fact that it’s having to invest heavily to remain competitive in several key areas has held margins down below where they might be.

    via Intel (PDF)

    ★ Samsung Reports Record Operating Margin Thanks to Semiconductor Strength (Apr 26, 2017)

    This content requires a subscription to Tech Narratives. Subscribe now by clicking on this link, or read more about subscriptions here.

    ARM-Based Windows 10 PCs to Arrive in Q4 2017 (Apr 21, 2017)

    This content requires a subscription to Tech Narratives. Subscribe now by clicking on this link, or read more about subscriptions here.

    Google’s Tensor Processing Unit Team Loses Key Members to Startup (Apr 21, 2017)

    This content requires a subscription to Tech Narratives. Subscribe now by clicking on this link, or read more about subscriptions here.

    Qualcomm Files Response to Apple Lawsuit (Apr 11, 2017)

    Qualcomm has now officially filed a response to Apple’s lawsuit over anticompetitive practices and breach of contract, including both answers to the specific allegations in the suit and a number of counter-claims. One of the main counterclaims is that, by “inducing” regulators to look into Qualcomm, Apple breached the companies’ “Cooperation agreement” and therefore was no longer entitled to certain payments it had received previously. The document further alleges that Apple made many false statements in the course of both its own suit and the discussions it had with regulators, and tried to insert itself into relationships between Qualcomm and other Apple suppliers. Perhaps most interestingly, Qualcomm brings to light something which was covered in the press at the time but didn’t get much attention: the allegation that Apple deliberately hamstrung the Qualcomm chips in the iPhone 7 such that performance would be consistent with those models that had Intel modems, and then prevented Qualcomm from talking about it. On the face of it, that allegation has nothing to do with the broader allegations, but it’s an area where Apple’s public reputation could be vulnerable, and I’m guessing it’s been included in the suit to garner more attention than Qualcomm would get through focusing on the patent and other issues alone.

    via Qualcomm (full document here (PDF))

    Google Shares Performance Characteristics for its Machine Learning Chip (Apr 5, 2017)

    It’s time to roll out that old Alan Kay maxim again: “those who are serious about software should make their own hardware”. Google started working on its own machine learning chip, which it calls a Tensor Processing Unit or TPU, a few years back, and has now shared some performance characteristics, suggesting that it’s more efficient and faster than CPUs and GPUs on the market today for machine learning tasks. While Nvidia and others have done very well out of selling GPU lines originally designed for computer graphics to companies doing machine learning work, Google is doing impressive work here too, and open sourcing the software framework it uses for machine learning. As I’ve said before, it’s extremely hard to definitively answer the question of who’s ahead in AI and machine learning, but Google consistently churns out evidence that it’s moving fast and doing very interesting things in the space.

    via Google Cloud Platform Blog

    Apple GPU Supplier Imagination Tech Says Apple Plans to Build its Own GPU in 1-2 Years (Apr 3, 2017)

    This already feels likely to be one of the biggest news items of the week (incidentally, you can now use the Like button below to vote for this post if you agree – the posts that get the most votes are more likely to be included in my News Roundup Podcast at the end of the week). There have been ongoing reports that Apple would like to build more of its own in-house technology, and GPUs have seemed at least a candidate given that Apple was said for a while to be mulling an acquisition of the company, and has been bringing Imagination Tech employees on board since the deal didn’t go ahead. The GPU obviously has a number of existing applications, but GPU technology has increasingly been used for AI and machine learning, so that’s an obvious future direction, along with Apple’s reported investment in AR. Apple’s ownership of its A-series chips (and increasingly other chips like its M and W series) is a key source of competitive advantage, and the deeper it gets into other chip categories, the more it’s likely to extend that advantage in these areas. This is, of course, also a unique example of Apple making a direct statement about a future strategy (albeit via a third party): as Apple is IMG’s largest customer, it had to disclose the guidance from Apple because it’s so material to its future prospects – the company’s share price has dropped 62% as of when I’m writing this.

    via Imagination Technologies

    Intel “Sets the Record Straight” on Moore’s Law (Mar 28, 2017)

    Intel is having a technology day today at which it’s talking up its manufacturing capabilities and offering evidence that not only is Moore’s Law alive and well but that Intel is ahead of its competitors in driving innovation in chip making. Intel obviously has a very strong incentive to argue that Moore’s Law isn’t – contrary to recent claims and reporting – dead, both as the originator of the idea and as perhaps its biggest corporate beneficiary. The reality is that we continue to see many of the benefits of Moore’s Law playing out, even if it isn’t necessarily playing out at exactly the same speed and in the same way as in the past (though Intel argues that it is still doing those things). Intel’s claim to be three years ahead of its competitors in process technology would certainly be disputed by major competitors and I suspect most analysts too – that’s hyperbole which isn’t supported by the reality. But Intel does lay out some interesting evidence to back its claim that there’s plenty of headroom left in Moore’s Law.

    via Intel

    Qualcomm blocked Samsung from selling Exynos chips: Korean Regulator – ZDNet (Mar 27, 2017)

    Another crazy wrinkle in the ongoing set of regulatory and legal actions against Qualcomm over anticompetitive practices: the Korean regulator responsible for the fine against Qualcomm last year says that one of the conditions of the contract between the companies was that Samsung would not be allowed to sell its own Exynos chips to any other vendors. What’s particularly crazy here is that Samsung is both Qualcomm’s biggest customer for chips and a contract manufacturer of those chips, so the two are inextricably intertwined here but are still going through this painful process. Samsung isn’t suing Qualcomm as Apple is, but it’s still likely cooperating with the authorities who are looking into its dealings in various markets. Just another sign of how far relationships between Qualcomm and some of its biggest customers have got, that they’re willing to start airing their grievances despite their close ties.

    via ZDNet

    Nvidia Partners with Bosch and Truck Maker Around Autonomy (Mar 16, 2017)

    Here are two partnership announcements from Nvidia, the first a deal with automotive component maker Bosch to incorporate Nvidia chips in self-driving solutions, and the second with truck maker PACCAR for self-driving truck technology. Nvidia continues to be one of the biggest names in autonomous driving, and certainly one of the most successful chip vendors (hence Intel’s Mobileye deal). These deals come on top of lots of existing ones, but trucks are a particularly interesting area – it feels like that’s a segment of the market that could actually see real-world adoption of autonomy much sooner than cars.

    via The Verge (Bosch) and Techcrunch (trucks)

    Intel in $15 Billion Deal for Self-Driving Tech Firm Mobileye – WSJ (Mar 13, 2017)

    It’s striking to me that this piece never mentions Nvidia once, even though that chip company has been making much of the noise in the automotive space over the past year, especially when it comes to autonomous driving. That prominence is clearly a driver for this deal, with Intel signing some of its own deals but not getting nearly the buzz Nvidia has been. Mobileye, meanwhile, has been striking deals left and right with a variety of players (though it recently ended its relationship with Tesla). Just in the last three months, it’s announced partnerships with HERE and BMW over mapping, but it’s also in many of other car manufacturers’ existing cars and their autonomous plans. Given Intel’s ongoing struggles in the mobile space, its recent loss of the Microsoft server business to Qualcomm and the ARM architecture, and the ongoing stagnation in PCs, it needs some new drivers of growth, and in-car technology could provide that. Mobileye is also attractive as a business – it’s growing fast (almost 50% year on year in 2016) and profitable, with fairly high margins. So this isn’t an acquisition that will take years to contribute to Intel’s business, although its overall scale is still small. But it’s integration opportunity and the eventual opportunity to sell a joined up solution of chips and sensors which Intel will be focusing on here.

    via WSJ

    Microsoft Pledges to Use ARM Server Chips, Threatening Intel’s Dominance – Bloomberg (Mar 8, 2017)

    The data center business at Intel accounts for almost a third of its revenues, has high margins, and has been growing considerably faster than its Client Computing segment (which includes PCs, tablets, and mobile phones). And it’s done well in large part because of commitments from big players like Microsoft to using its chips in their data center servers. But now Microsoft is saying it plans to switch to using ARM-based chips made by Qualcomm in its Azure server infrastructure instead, which could put a dent in Intel’s future growth and reduce its share from the 99% cited in this Bloomberg article. This isn’t imminent – it’s a step on a path Microsoft is committed to, but hasn’t been rolled out to any customer facing servers yet. But ARM-based chips have been cited as potential substitutes for Intel chips in server farms for some time now, so this could be the beginning of a dramatic shift in the next few years. That’s obviously terrible news for Intel, for which the data center business has been a useful source of growth and margins in recent years. Meanwhile, this is such a small business for Qualcomm today that it doesn’t even get mentioned in its quarterly earnings materials, but that could obviously change rapidly going forward.

    via Bloomberg

    Xiaomi Launches Its Own Chip, With an Assist From Beijing – WSJ (Feb 28, 2017)

    I commented on the reports a couple of weeks back that Xiaomi would be building its own chips, and guessed that Xiaomi would likely start at the low end of its device range and work up from there, and that’s exactly what they’re doing: the Mi 5C is the first phone using Xiaomi’s homegrown chips, and sells for a little over $200. It’ll be interesting to see what if anything comes out of the reviews of the phone about its performance relative to Xiaomi’s earlier low-end phones – a solid early performance is critical for building confidence that Xiaomi knows what it’s doing here. The company also said it had spent a billion yuan – around $145 million – on building its capability, and that it received some help from the Chinese government, though it’s not clear how much. To put that in context, Apple’s acquisition of PA Semi alone cost $278 million, and that’s before all the additional work and money it put in organically following the acquisition to build its own chips. So though Xiaomi is splashing out somewhat here, it’s still a small investment in the context of earlier similar investments.

    via WSJ

    Intel’s 7560 Modem Could Push Next iPhone to 1Gbps – PCMag (Feb 21, 2017)

    There’s some conjecture here on two points: that simply because Intel has added CDMA/EVDO capability to its next LTE modem Apple will use it globally, and that a modem capable of delivering peak throughput of a gigabit per second will actually do so in real world environments. The latter is an obvious stretch, given that real-world performance is always a fraction of the theoretical peak, but the former may well be a stretch too. I’m not convinced that Intel could ramp up production quickly enough to be the supplier for all of the next generation of iPhones – that would be a massive step up over its iPhone 7 supply. And I’m not convinced that Apple, having finally gained a measure of redundancy by dual sourcing modems for the iPhone 7, would so quickly jump back into single sourcing, especially given the performance limitations of the current generation Intel modems. That’s not to say this would never happen, and it’s obviously a very interesting point of leverage in the context of the current bad blood between Apple and Qualcomm, but I still think it’s somewhat far fetched at this point.

    via PCMag

    Why Apple Is One Of The Most Innovative Companies Of 2017 – Fast Company (Feb 16, 2017)

    I’ve never been a big fan of these kinds of ranking exercises – they’re often arbitrary, designed to make news as much as come up with the right answer, and basically meaningless in the real world. So why am I including this item here? It’s about the specific reason why Apple was chosen here: its silicon chips. This is so overlooked as a source of leadership and differentiation for Apple and yet it’s absolutely critical. It took the acquisition of PA Semi and lots of other work besides, but Apple has spent years perfecting its chip design and it now pays off in massive ways for the company. No wonder you see Xiaomi and others pursuing their own chip strategies, but no-one should assume that’s a straightforward or quick process.

    via Fast Company (you might also be interested in this episode of the Beyond Devices Podcast, in which we did a deep dive on Apple’s A-series chips)

    Apple Developing New Mac Chip in Test of Intel Independence – Bloomberg (Feb 1, 2017)

    The first ARM-based chip in a Mac arrived with the new MacBook Pro with Touch Bar late last year, but it handled a brand-new function. Bloomberg is reporting that another ARM-based chip could take over functionality currently performed by the main Intel chip, which would be something of a first. Specifically, it’s the Power Nap function in macOS which the ARM chip would handle, something particularly well suited to a low-power, energy efficient chip. Naturally, all of this also enables Apple to continue testing ARM-based chips in Macs as a way of seeing which functions they might take over from Intel over time, with a possible endgame of replacing Intel entirely. Apple has been somewhat frustrated in recent years by its dependence on Intel’s roadmaps and the way PC makers have benefited more from recent upgrades to its chips than Apple has. Reducing its dependence on Intel in its Mac lineup might give Apple more freedom to move faster and differentiate versus PCs, and would also put pressure on Intel to help Apple do the same with its own chips.

    via Bloomberg

    Source: Google’s Pixel 2 to feature improved camera, CPU, higher price, but ‘budget’ Pixel also in works | 9to5Google (Jan 26, 2017)

    Lots of interesting stuff in here, but of course none of it certain to pan out in the actual product. To my mind one of the most interesting aspects is that the price of the Pixel 2 is expected to be $50 higher than the first-generation Pixel, which was already priced at premium levels – that part doesn’t ring quite true to me, unless it’s an Apple-like creation of an additional tier above the standard ones. Pricing its only phone higher than the base iPhone model would unnecessarily limit the market, and that seems unlikely. The camera focus makes sense – the first version majored on the camera and the Assistant differentiator should have been eroded relative to other Android devices by then, so hardware features will be important. The budget version is also interesting in the context of the recent reports about Google bringing Android One to the US – it certainly wouldn’t be unprecedented for two parts of Google to be working on essentially the same problem in different ways, but the hardware strategy there has been more joined up lately. The other thing to note is the details about chips, as Google is apparently testing both Samsung and Intel components as well as something it’s built itself. There have been repeated rumors about Google building ARM-based servers, and it’s possible that it’s also experimenting with its own ARM-based chips for smartphones too, though this would be a massive multi-year effort, especially tough without an acquisition of significant existing skills a la Apple/PA Semi.

    via 9to5Google