Posts Tagged‘arm’

The One Horse Race That is CPUs.

Roll back the clock a decade or so and the competition for what kind of processor ended up in your PC was at a fever pitch with industry heavyweights Intel and AMD going blow for blow. The choice of CPU, at least for me and my enthusiast brethren, almost always came down to what was fastest but the lines were often blurry enough that brand loyalty was worth more than a few FPS here or there. For the longest time I was an AMD fan, sticking stalwartly to their CPUs which provided me with the same amount of grunt as their Intel brethren for a fraction of the cost. However over time the gap between what an AMD CPU could provide and what Intel offered was too wide to ignore, and it’s only been getting wider since then.

AMD Logo Official

The rift is seen in adoption rates across all products that make use of modern CPUs with Intel dominating nearly any sector that you find them in. When Intel first retook the crown all those years ago the reasons were clear, Intel just performed well enough to justify the cost, however as time went on it seemed like AMD was willing to let that gap continue to grow. Indeed if you look at them from a pure technology basis they’re stuck about 2 generations behind where Intel is today with the vast majority of their products being produced on a 28nm process, with Intel’s latest release coming out on 14nm. Whilst they pulled a major coup in winning over all of the 3 major consoles that success has had much onflow to the rest of the business. Indeed since they’ll be producing the exact same chips for the next 5+ years for those consoles they can’t really do much with them anyway and I doubt they’d invest in a new foundry process unless Microsoft or Sony asked them nicely.

What this has translated into is a monopoly by default, one where Intel maintains it’s massive market share without having to worry about any upstarts rocking their boat. Thankfully the demands of the industry are pressure enough to keep them innovating at the rapid pace they set way back when AMD was still biting at their heels but there’s a dangerously real chance that they could just end up doing the opposite. It’s a little unfair to put the burden on AMD to keep Intel honest however it’s hard to think of another company who has the required pedigree and experience to be the major competition to their platform.

The industry is looking towards ARM as being the big competition for Intel’s x86 platform although, honestly, they’re really not in the same market. Sure nearly every phone under the sun is now powered by some variant of the ARM architecture however when it comes to consumer or enterprise compute you’d be struggling to find anything that runs on it. There is going to have to be an extremely compelling reason for everyone to want to translate to that platform and, as it stands right now, mobile and low power are the only places where it really fits. For ARM to really start eating Intel’s lunch it’d need to make some serious inroads into those spaces, something which I don’t see happening for decades at least.

There is some light in the form of Kaveri however it’s less than stellar performance when compared to Intel’s less tightly coupled solution does leave a lot to be desired. At a high level the architecture does feel like the future of all computing, well excluding radical paradigm shifts like HP’s The Machine (which is still vaporware at this point),  but until it equals the performance of discreet components it’s not going anywhere fast. I get the feeling that if AMD had kept up with Intel’s die shrinks Kaveri would be looking a lot more attractive than it is currently, but who knows what it might have cost them to get to that stage.

In any other industry you’d see this kind of situation as one that was ripe for disruption however the capital intensive nature, plus an industry leader who isn’t resting on their laurels, means that there are few who can hold a candle to Intel. The net positive out of all of this is that we as consumers aren’t suffering however we’ve all seen what happens when a company remains at the top for far too long. Hopefully the numerous different sectors which Intel is currently competing in will be enough to offset their monopolistic nature in the CPU market but that doesn’t mean more competition in that space isn’t welcome.

The Windows 8/RT Distinction is Clear, Should You Not be an Idiot.

I’ve been using Windows 8 as my main system for the better part of 2 months now and, whilst I’ll leave the in-depth impressions for the proper review, I have to say I’m pretty happy with it. Sure I wasn’t particularly happy with the way things were laid out initially but for the most part if you just blunder along like its Windows 7 you’re not going to struggle with it for very long. I might not use any of the modern styled applications, they don’t feel like they’re particularly well suited to the mouse/keyboard scenario if I’m honest, but everything else about it works as expected. Of course whilst Microsoft has already sold 40 million licenses of Windows 8 most people are focusing on Windows RT, care of the Surface tablet.

For the technically inclined the differences between the two are pretty stark and we’ve known for a long time that the Surface is essentially Microsoft’s answer to the iPad. The lines are a little bit more blurry between Surface/RT and the full version of Windows 8, thanks to the Modern Styled UI being shared between them, but the lack of a desktop made it pretty clear where the delineation lay. It seems however that there’s a feeling among some the bigger media outlets that Windows 8 is suffering from an identity crisis of sorts which has been perplexing me all morning:

What we’re seeing, I think, is Microsoft dancing around an uncomfortable reality: Windows RT just doesn’t have much to offer, so it’s hard to explain how it’s different from Windows 8 without making it look inferior.

The only distinct advantage for Windows RT is its support for “connected standby,” a power-saving mode that lets the device keep an eye on e-mail and other apps while it’s not in use. It’s a nice feature to have, but on its own it’s a tough sell compared to Windows 8′s wider software support. (UPDATE: As Eddie Yasi points out in the comments, the Atom-based chips that Windows 8 tablets are using, codenamed Clover Trail, support connected standby as well.)

The main thrust of the article, and another one it linked to, is that there’s been no real information from Microsoft about the differences between the fully fledged version of Windows 8 and its RT cousin. I’ll be fair to the article and not use anything past its publication date but for anyone so inclined I wrote about the differences between the two platforms well over a year ago and I was kind of late to the party on it too. Indeed the vast majority of the tech press surrounding the Surface release understood these differences quite clearly and it appears that both Time and The Verge were both being willingly ignorant simply to get a story.

Granted The Verge has something of a point that the retail representatives didn’t know the product but then again why were you asking in depth technical questions of a low wage retail worker? Most people who are looking for a Surface/iPad like device aren’t going to want to know if their legacy applications will run on it because, to them, they’re not the same thing. You could argue that the customer might have seen the Modern UI at home and then assumed that the Surface was exactly the same but I’d struggle to find someone who had installed Windows 8 this early in the piece and wasn’t aware that the Surface was a completely different beast.

Indeed the quote paragraphs above imply that Jared Newman (writer of the Time article) isn’t aware that the RT framework, the thing that powers the Modern  UI, is the glue that will join all of Microsoft’s ecosystem together. Not only does it underpin Windows 8 but it’s also the framework for Windows Phone 8 and (I am speculating here but the writing is on the wall) the upcoming Xbox. What Windows RT devices offer you is the same experience that you’ll be able to get anywhere else Microsoft ecosystem but on low power devices. Newman makes the point that they could very well run them on Atom processors however anyone who’s actually used one can tell you that their performance is not up to scratch with their i3/5/7 line and is barely usable for desktop applications. They’re comparable in the low power space, meaning they would have made a decent replacement for ARM, but considering that 95% of the world’s portable devices run on the ARM it makes much more sense to go with the dominant platform rather than using something that’s guaranteed to give a sub-par experience.

I don’t like doing these kinds of take down posts, it usually feels like I’m shouting at a brick wall, but when there’s a fundamental lack of understanding or wilful ignorance of the facts I feel compelled to say something. The Windows8/RT distinctions are clear and, should you do even a small amount of research, the motives for doing so are also obvious. Thankfully most of the tech press was immune to this (although TechCrunch got swept up in this as well, tsk tsk) so there’s only a few bad apples that needed cleaning up.

Apple Eyeing ARM For Their Desktop Line.

The name of the game for all large technology companies is platform unification and domination, with each of them vying to become the platform that consumers choose. Microsoft has been on a long and winding road to accomplishing this since they first talked about it 3 years ago and Apple has been flirting with the idea ever since it started developing its iOS line of products with features like the App Store making its way back into OSX. Neither of them are really there yet as Windows 8/WinRT are still nascent and requiring a lot more application development before the platform can be considered unified and there is still a wide schism betwen iOS and OSX that Apple hasn’t really tried to bridge.

Predominately that’s because Apple understands that they’re two different markets and their current product strategy doesn’t really support bridging those two markets. The iOS space is pretty much a consumer playground as whilst you can do some of the things that Apple’s previous OS was known for on there its far from being the creative platform that OSX was (and still is, to some extent). Indeed attempts to bridge their previous products with more consumer orientated versions have been met with heavy criticism from their long time fans and their failure to provide meaningful product updates to their creative powerhouse the Mac Pro has also drawn the ire of many creative professionals.

If I’m honest I didn’t really think that Apple would turn their backs on the creative niche that is arguably responsible for making them what they are today. It’s understandable from the company’s point of view to focus your attention on the most profitable sectors, much like games developers do with the whole console priority thing, but it almost feels like the time when Apple still considered itself a player in the enterprise space, only to quietly withdraw from it over the course of a couple years. Whilst there isn’t much evidence to support this idea the latest rumours circulating that they may be considering a switch to ARM for their desktop line doesn’t help to dispel that idea.

ARM, for the uninitiated, is a processor company based out of Cambridge that’s responsible for approximately 95% of all the processors that power today’s smartphones. They are unquestionably the kings of the low power space with many of their designs being able to achieve incredible efficiencies which is what enables your phone to run for hours instead of minutes. Whilst they may no longer be the supplier for the chips that powers Apple’s current line of iOS products their technology is still the basis for them. Suffice to say if you’ve got any piece of mobile technology it’s likely that there’s some kind of ARM processor in there and it’s the reason why Microsoft chose it as their second platform for the WinRT framework.

Apple switching platforms is nothing new as they made the switch to x86/Intel back in 2006. The reason back then was that PowerPC, made by IBM, was not able to keep pace with the rapid improvements in performance that Intel was making but was also because of the performance-per-watt of their processors which was arguably why AMD wasn’t considered. Apple’s direction has changed considerably since then and their focus is much more squarely aimed at portable experiences which is far better served by the low power processors that ARM can deliver. For things like the MacBook and the Air lower power means a longer battery life, probably the most key metric by which these portable computers are judged by.

There’s no doubt that Apple will be able to make the transition however I’m not sure that the cost to them, both in real and intangible terms, would be worth it. Forgetting all the technical challenges in getting all your third parties to re-architect their applications the unfortunate fact is that ARM doesn’t have a processor that’s capable of performing at the same level that Intel’s current line is. This means for creative applications like photo/video editing, graphic design and the like their current software suites will simply not be viable on the ARM platform. Since the transition is a ways off its possible that ARM might be able to design some kind of high power variant to satisfy this part of the market but traditionally that’s not their focus and since the desktop sector is one of Apple’s smallest revenue generators I can’t see them wanting to bother doing so.

This is not to say that this would be a bad move for Apple at large however. Being able to have a consistent architecture across their entire line of products is something that no other company would be able to achieve and would be an absolute boon to those seeking a ubiquitous experience across all their devices. It would also be a developer’s wet dream as you could make a cross-platform applications far more easily than you could with other platforms. Considering that Apple makes the majority of its money from ARM based platforms it doesn’t come as much surprise that they might be considering a move to it, even if that’s at the cost of creative sector that brought them back from the graveyard all those years ago.

I don’t usually comment on Apple rumours, mostly because they’re usually just a repeat of the same thing over and over again, but this one caught my attention because if it turns out to be true it will mark Apple’s final step away from its roots. Whilst the creative professionals may lament the disappearance of a platform they’ve been using for over 2 decades the saving grace will be the fact that on a feature level the Windows equivalents of all their programs are at feature parity. Apple will then continue down the consumer electronics road that it has been for the past 10+ years and where it will go from there will be anyone’s guess.

Bring Your Own Device: The Line That Microsoft Has Drawn in the Sand.

I’ll be honest the bring your own device movement annoys the hell out of me as an IT administrator. I think this is mostly because the movement starts from higher up, usually when an executive discovers how wonderful it is to read personal email on his iPad and then wants the same thing for work. Queue a rushed, short term project that involves putting in all manner of hacks, poorly documented systems and as of yet unvetted devices being introduced into the network. I guess if you read inbetween the lines on that one I don’t really have a problem with the BYOD movement per se, just the way it’s weaselled its way into the environments I’ve been responsible for.

That being said I’m not one to stand in the way of inevitable change and every day it’s looking more and more like the BYOD movement is something that I’d rather embrace than struggle against. It’s still a nascent movement, with all the associated problems, but thankfully we have many companies that are taking notice of this movement and ensuring that these devices can be integrated seamlessly into corporate environments. The next version of Windows has some provisions in it for supporting BYOD but there’s an interesting delineation between those devices and your traditional corporate computing device.

Windows 8 brings with it a new control panel option that allows users to connect to the corporate network using their email address and a password. Once they’re authenticated their device then downloads a series of approved apps from the corporate network like the one shown in the picture above. You can also provide access to applications in the Microsoft Marketplace through an on-site cache. What’s missing here however is any control over the end device; you can’t enforce things like a password policy or on-device encryption should you use this method. Additionally Windows 8 devices on the ARM architecture are not able to be members of an Active Directory domain, a critical feature for most large enterprises.

What this means is that Microsoft, whilst embracing the BYOD movement with one hand, is drawing a clear line in the sand between where traditional corporate computing resources lie and what untrusted and unvetted have access to. It may seem like an odd line to draw as you’re basically relegating BYOD users to be second class citizens on your network but in reality granting users the ability to control the platform means you can’t trust it in the same way you trust something that’s under your control. This is probably the most happy compromise that Microsoft could come up with and to be honest it’s actually not that bad.

This kind of interoperability between unknown Windows 8 devices and trusted networks provides a lot of opportunities for innovation in the corporate app space. The applications delivered with the initial app package can be highly tailored towards a streamlined user experience, one that could be unique to the user’s requirements. Take for example the HR app, you could have different versions for HR staff, management and end users all available through the access portal. Reworking the interface to be friendly to these (most likely) touch centric devices would go a long way to improving the current state of corporate applications which most users loathe to use.

Microsoft had to draw the line somewhere and realistically I’m surprised at the level of functionality that they’re granting BYOD users. The traditional approach has been to provide a secure container on top of the device and then enabling full access to the corporate environment. Whilst this works in theory Windows 8, especially on ARM devices, was designed with a different user interface paradigm in mind, one that centers around user experience rather than iterating on the current desktop. Corporations will have to embrace this if they want to take BYOD seriously and I believe that those who don’t will have their (rather irate) users to contend with.

Windows 8: First Step to the Realization of Three Screens.

The last two years have seen a major shake up in the personal computing industry. Whilst I’m loathed to admit it Apple was the one leading the charge here, redefining the smart phone space and changing the way many people did the majority of their computing by creating the wildly successful niche of curated computing (read: tablets). It is then inevitable that many subsequent innovations from rival companies are seen as reactions to Apple’s advances, even if the steps that company is taking are towards a much larger and broader goal than competing in the same market.

I am, of course, referring to Microsoft’s Windows 8 which was just demoed recently.

There’s been quite a bit of news about the upcoming release of Windows 8 with many leaked screenshots and even leaked builds that gave us a lot of insight into what we can expect of the next version of Windows. For the most part the updates didn’t seem like anything revolutionary although things like portable desktops and a more integrated web experienced were looking pretty slick. Still Windows 7 was far from being revolutionary either but the evolution from Vista was more than enough to convince people that Microsoft was back on the right track and the adoption rates reflect that.

However the biggest shift that is coming with Windows 8 was known long before it was demoed: Windows 8 will run on ARM and other System on a Chip (SOC) devices. It’s a massive deviation from Microsoft’s current platform which is wholly x86/x86-64 based and this confirms Microsoft’s intentions to bring their full Windows experience to tablet and other low power/portable devices. The recent demo of the new operating system confirmed this with Windows 8 having both a traditional desktop interface that we’re all familiar with and also a more finger friendly version that takes all of its design cues from the Metro interface seen on all Windows Phone 7 devices.

The differences between these two interfaces just don’t stop at what input device they were optimized for either. Whilst all Windows 8 devices will be capable running the huge back catalog of software that has been developed for Windows over the past few decades in the traditional desktop interface mode the new tablet optimized interface relies on applications built using HTML5 and JavaScript. This is arguably done so that they are much more platform independent than their traditional Windows applications cousins who, whilst most likely being able to run since .NET will be ported to the ARM and SOC infrastructures, won’t have been designed for the tablet environment. They’ll still be usable in a pinch of course but you’d still want to rewrite them if a large number of your users were moving to the tablet/smartphone platform.

Looking at all these changes you can’t help but think that they were all done in reaction to Apple’s dominance of the tablet space with their iPad. It’s true that a lot of the innovations Microsoft has done with Windows 8 mirror those of what Apple has achieved in the past year or so however since Windows 8 has been in development for much longer than that not all of them can be credited to Microsoft playing the me-too game. Realistically it’s far more likely that many of these innovations are Microsoft’s first serious attempts at realizing their three screens vision and many of the changes in Windows 8 support this idea.

A lot of critics think the idea of bringing a desktop OS to a tablet form factor is doomed for failure. The evidence to support that view is strong too since Windows 7 (and any other OS for that matter) tablet hasn’t enjoyed even a percentage of the success that the dedicated tablet OS’s have. However I don’t believe that Microsoft is simply making a play for the tablet market with Windows 8, what they’re really doing is providing a framework for building user experiences that remain consistent across platforms. The idea of being capable of completing any task whether you’re on your phone, TV or dedicated computing device (which can be a tablet) is what is driving Microsoft to develop Windows 8 they way they are. Windows Phone 7 was their first steps into this arena and their UI has been widely praised for its usability and design and Microsoft’s commitment to using it on Windows 8 shows that they are trying to blur the lines that current exist between the three screens. The potential for .NET applications to run on x86, ARM and other SOC platforms seals the deal, there is little doubt that Microsoft is working towards a ubiquitous computing platform.

Microsoft’s execution of this plan is going to be vital for their continued success. Whilst they still dominate the desktop market it’s being ever so slowly eroded away by the bevy of curated computing platforms that do everything users need them to do and nothing more. We’re still a long time away from everyone out right replacing all their PCs with tablets and smart phones but the writing is on the wall for a sea change in the way we all do our computing. Windows 8 is shaping up to be Microsoft’s way of re-establishing themselves as the tech giant to beat and I’m sure the next year is going to be extremely interesting for fans and foes alike.