Slashdot Mirror
Intel Announces Major Reorg To Combine Mobile and PC Divisions
MojoKid writes: For the past year, Intel has pursued what's known as a "contra-revenue" strategy in its mobile division, where product is deliberately sold at a loss to win market share and compete effectively. This has led to a huge rise in tablet shipments, but heavy losses inside Intel's mobile division. Today, the company announced that it would take steps to fold its mobile and conventional processors into a single operating division. While this helps shield the mobile segment from poor short-term results, it also reflects the reality that computing is something users now do across a wide range of devices and multiple operating systems. Intel may not have hit anything like the mobile targets it set out years ago, but long-term success in laptops, tablets, and smartphones remains integral to the company's finances. Desktops and conventional laptops are just one way people compute today and Intel needs to make certain it has a robust long-term presence in every major computing market.
where product is deliberately sold at a loss to win market share and compete effectively
Isn't that also called dumping?
Seems like they want to conceal how well/poorly they are doing in the mobile sector. Makes a certain amount of sense most people see the mobile market as the hot new future.
Given the fairly lame update to the Mac Mini caused mainly by the lack of choices in Intel's mobile CPU offerings (and Apple's refusal to design and stock a separate motherboard just for quad core), I'm wondering just what would it take for Apple to make yet another CPU transition. They must hate being dependent on the release schedules of Intel for when it comes to putting out Macs, and the A8X is nearly the performance of a couple years ago MacBook Air.
Phones, tablets, laptops, all is mobile. The days of tower rigs are over.
Given that a 'tower rig' is basically a server turned on its side, with fewer 40mm fans and some of the classy reliability features cut, that category will take a great deal of killing. On the other hand, the CPU in a server or tower is almost certainly using nearly as many of the power gating, adjustable clock speed, and various other thermal protection and power saving strategies as the mobile CPUs are. Overall efficiency is still going to be lower ('eh, we're on AC, just keep the PSU energized so a USB peripheral can wake the system!' isn't god's gift to brilliant standbye power numbers); but 'mobile' and 'desktop' have been on something of a collision course ever since the P4 flamed out, almost literally, and Pentium M derivatives took over.
This is Intel basically admitting defeat in the mobile space. It's good they don't feel so cocky anymore. Competition is good for everyone.
That might be marginally true, but only for consumers. The people who actually make the shit everyone consumes still need those tower rigs. Try running MSVC, Photoshop, Max, CRYENGINE and other DAW tools, all at once on your phone or tablet. You could run it on a beefy laptop, but that would seriously cost several times more than a standard PC and still be worse at the job.
So sure, over for the people who's only real need is a web browser...still going strong for the rest of us.
All those moments will be lost in time, like tears in rain.
Been used effectively for years.
Intel is the world leader when it comes to silicon advancement, there's no doubt of that. Their quite real and appropriately named 14nm process starts shipping incredibly soon, and TSMC/Global Foundries, now seemingly their only competition, don't even hope to have their inappropriately named "16nm" process shipping in products for two years out, a process not actually much to any denser than the just now shipping TSMC 20nm process.
Intel's Core technology is also excellent in terms of silicon to performance. The problem comes in at the cost of that world leading silicon. TSMC has concentrated on costs, and while 20nm process might be a day late and a few dollars short in terms of performance and density, in terms of cost TSMC can make a profit at the same price Intel can produce chips at, let alone sell them.
And with mobile products being relatively cheap, and their prices coming down for the most part, its that profit at a low cost to consumers and high volume that's selling. Intels Atom products are actually perfectly competitive for performance. They just cost Intel too much to produce at their super fancy fabs and that just have to be cutting edge instead of cost efficient.
This. My core i7 laptop with AMD graphics and maxed out on memory, is only marginal for use doing electronic design, eCAD and PCB layout. It's OK for software compiles; but, all my projects target smallish embedded processors. Time is money... If I have to wait for something then I'm pissing away money. I have the laptop because I needed a portable machine to carry with me to customer sites.
Isn't that also called dumping?
Strictly speaking no it is not dumping. Dumping is the act of charging less in a foreign market than you charge in your domestic market. That isn't what Intel is doing. What Intel is doing might be considered a form of predatory pricing but it isn't dumping. All dumping is predatory pricing but not all predatory pricing is dumping.
Surely using their market power in one segment to sell at a loss in a different segment is anti-competitive.
SURELY NOT!!!!!
Intel needs to get behind an organized effort to bring us a business-grade mobile device. That's the only low hanging fruit left. Take the following excerpt I pulled from an article:
Let's rewind to 2007. RIM owns the mobile space for business, while consumer devices are primarily "dumb phones". In comes Apple, flush with iPod money, and looking for the next evolution of it's highly profitable device. The solution is simple: why carry an iPod and a phone? Thus, the iPhone is born.
In a single generation the iPhone brings massive innovation to the market. The device is targeted at Apple's primary demographic: the consumer, but the features are so beyond what is currently available that this type of smartphone doesn't take long to become a favorite in the business commnuity as well.
The large touch screen destroys the conventional track ball/pad, allowing the user to display more text, and use multi-touch to navigate more efficiently. The full webkit browser completely destroys the WAP-based dinosaurs giving the user a desktop grade browser at their finger tips. The user can carry all forms of media with them and display it at their whim. And, finally, and most importantly, the design of the operating system is centered around a robust API which doesn't take long to bloom into a wealth of independent applications that let the user do things they never before thought possible.
The response at RIM is unforunately short-sighted. RIM sees the device as a "toy". It sees it as a consumer-grade flash in the pan that will eventually collapse in the face of the established security and familiarity of their Enterprise Server platform, and BBM. RIM does opt to borrow some of the innovations - like the touch screen - and implement it their own, poorly advised ways but, ultimately, things at RIM continue as usual.
Now let's fast forward to 2013. The market has spoken. Blackberry market share is down to single digits and the company needs to do something quick to turn things around. They've been working for years on something that is supposed to change our lives and we're finally going to get to see it. What they unveil is astonishing: a consumer-targeted device.
The playing field in consumer-grade devices is now beyond saturated. We've had Google, Apple, and even Microsoft all battling each-other for the last 5 years. Innovation year-over-year is staggering. Why blackberry decided to try to compete in this market is baffling. What's worse, is they released an inferior product, on their own independent platform, that - of course - is going to gather no developer support in an already saturated market.
So here we are, 2014 and - still - no business-grade device in the mobile market. We have a dizzying amount of consumer-grade choice, but nothing properly designed with business in mind. In response I would like to say the following to the entire tech community involved in mobile device development:
We're here. We have money. We have a lot more money than all these teenage kids. Please, please, I want to spend it. Someone give me a business-grade mobile phone and tablet. Important things to me are: checking my email, security, centralized device management, and integration with existing business technologies. Reward: see Microsoft's stock price in the 90's.
If it ain't broke, don't fix it.
Phones, tablets, laptops, all is mobile. The days of tower rigs are over.
From an architectural POV, I wouldn't bundle laptops w/ phones/tablets. ChromeOS is still a fringe OS, and Linux/BSD even more so. Laptops are still overwhelmingly Wintel, and even WinRT failed to make any impact. You might as well bunch laptops w/ your 'tower rigs'.
ARM pretty much owns the phone & tablet markets with not just Android, but iOS as well. Intel pretty much owns the laptop and server markets w/ Wintel. A good indicator is Apple, who've gone w/ ARM for their iOS toys - iPhones & iPads, while they've gone w/ Intel on all their OS-X boxes.
The only 'mobile' area that Intel could do well is the Surface Pro clones, once it's recognized that it's a de-facto laptop, rather than a tablet market that they are playing in. There is no reason to go Intel for Android.
Most of Intel's desktop Core i3s are STILL running circles around AMD in single threaded performance tests. Sure, if you can use 8+threads AMD has a few decent options, but that's not most consumer workloads.
Most consumer workloads won't tax either processor, the AMD chip costs a lot less, and the AMD chipset costs a lot less, too. When the system is "good enough" (even my old-ass 1045T is peppy both when puttering around and also when the system is heavily loaded) and literally a couple hundred dollars cheaper between the cost of the CPU and the cost of the motherboard, a lot of people are going to go AMD. I have other places to spend my money, and wringing a few more FPS out of a game isn't worth adding 20% or more to the cost of the system.
For highly-parallelized tasks, AMD is still cheaper than intel flop for flop, and if you're planning to throw all your servers away every few years as many businesses do, you can punt on the power consumption issue. For a home user, it's usually not even on the radar. And it's not like AMD is just burning power, either. The lowest-end systems still have better graphics than intel; though I'm certainly no fan of ATI graphics, intel is only now getting serious about graphics performance, or perhaps that's becoming competent in.
If you need/want balls-out single thread performance, or can be convinced that it's important to you even when it isn't, sure you're going to buy intel. But you're going to pay a premium. It has been ever thus. At times, it made great sense, because for example around the P55C vs. K6 days everyone else had apparently forgotten completely how to make a chipset. Today, not so much.
"You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
I'm not trying to be a troll or flame here, so please don't take it like that. But, you may be comparing AMD processors and Intel in the same price range. Because the highest passmark bench results for laptops certainly belong to the 4th generation Intel mobile processors.
You must also work in an area of the country that is either really ahead, or really behind. I'm Sr sysadmin for a medium sized company and I haven't encountered a single person - outside of a geologist or engineer that needs real power - who prefers a desktop to a laptop in many years.
If it ain't broke, don't fix it.
The problem with your post is that on newegg, 48 USD gets you an AMD A6-5400K while 46 USD gets you an Intel Celeron G1610.
The Celeron is actually a bit faster faster CPU and uses less power, although the A6 has a much faster GPU.
Intel is, essentially, in the enviable position of having chips which are faster, consume less power and are actually smaller, thus cheaper to manufacture.
And they are segmenting and slicing the market as they wish.
Yes, they offer top notch performance at a premium. And even more performance if you pay an arm and a leg.
But if you just want something cheap and competent, they got that too.
So does this mean they will be releasing a processor that runs applications equally as poor on mobile and desktop? They will probably straighten in out in Intel i8.1 or just skip to i10 where they re-introduce the PCI bus bridge they disabled in i8.
When it comes to building a compact PC, it's pretty hard to say they should have standardized on a socket which is only useable for 47W parts, versus a socket that supports parts which center around 15W. Are you actually saying that they should have designed a Mini with a 47W CPU? For home theatres and light server use? I happen to love the low energy use of the Mac Mini; I can't buy one for desktop use as I need the quad core for Xcode development, but I respect the idea of making a quiet, low energy general purpose computer.
I'm Sr sysadmin for a medium sized company and I haven't encountered a single person - outside of a geologist or engineer that needs real power - who prefers a desktop to a laptop in many years.
CPU power is not the only issue.
Desktops can be made really quiet, far more so than laptops of comparable power. There are a number of clever solutions to achieve this, each addressing different aspects of computer noise, such as the huge "NoFan" heatsinks. The lack of machine noise can be very useful in many work environments. Artists, for example, often value silence when they work.
Also, desktops can run really high end graphics cards: nice if you want good performance while running multiple big monitors (i.e. 30"). I'd really hate to develop software on even the largest laptop screen, after getting used to the convenience of the large desktop screens. One doesn't have to be a power-user (or be doing work requiring a high end CPU) to appreciate the large screens. A high end laptop might be able to drive these screens, but probably not very well.
Some jobs generate or work with lots of data, without necessarily requiring a high end CPU (i.e. data access speed, not CPU speed, limits throughput). It's much easier (and quieter) to cool a multi-hard-drive array in a big case, leading to longer drive life and better performance. Some of the better cases have 200mm fans, which can spin a lot slower than smaller fans while generating far better cooling. Laptops can't even come close to this, either in terms of cooling or drive size. The case for the array doesn't necessarily have to be the same case as one's motherboard, of course. But if you need the big case anyway, it might make sense to put the motherboard in it (and you'll probably have fewer issues with the RAID controller).
Physical security with desktops is much easier to enforce than with laptops, for those sites with stringent security requirements.
From a home use perspective, you won't get good quality graphics on multiple screens using a flight simulator, without a big case and real full-size graphics cards - no laptop can do this. For that matter, you won't get good quality graphics on even single screens for high end games without a really high end laptop (and one that isn't all that portable), which makes the (much less expensive) desktop alternative pretty appealing.
Desktops are still far more upgradable than laptops. A good case and power supply can outlast many motherboards, and give a net cost savings compared to purchasing multiple laptops.
Laptops tend to fail more often than desktops, and are harder to repair.
In my experience, the overhead types like the managers and marketing people prefer the laptops for convenience, but the savvy technical folks often still like the desktops.