Mobile devices: fast, light, long battery life… pick any two

Yesterday Microsoft announced their "Surface" line of mobile devices. I wrote that one of the problems with it is that it attempts to combine the best features of a tablet and a laptop, but in doing so it actually creates something that's the worst of both worlds.

Yesterday Microsoft announced their "Surface" line of mobile devices. I wrote that one of the problems with it is that it attempts to combine the best features of a tablet and a laptop, but in doing so it actually creates something that's the worst of both worlds. (A simplified explanation is that tablets are for consuming, laptops are for creating. Tablets don't replace laptops—they augment them.)

I was trying to explain my views in the comments of that article, and it occurred to me that we can build another triangle here.

You all know the old adage, "Fast, cheap, easy—pick two." (And you may remember the one I created three years ago about remoting protocols: low bandwidth, good user experience, low CPU—pick two.) Well today I'd like to introduce one that related to mobile devices like tablets and laptops:

Fast, light, long battery life—pick two.

Mobile device triangle

If you want a mobile device that's both fast (high performance processor) and light, then you can't expect long battery life. If you want a device that's fast and has long battery life, then it's not going to be light (since it will have to be full of batteries to support that fast performance). And if you want a device that's light and that has a long battery life, then it's not going to be fast.

This is an immutable law of physics that applies to all mobile devices. Sure, better technology and Moore's Law will mean that the quantitative definition of each of these will constantly move forward. And "fast" might mean different things to different people. (One person might argue that the dual core A5 in an iPad is "fast," but it's nothing compared to a quad core Ivy Bridge. But that's fine. Apple chose the "light" and "long battery life" options instead.)

Let's apply this to real life to see it in action (and to understand why I created it in the first place). I mentioned yesterday's announcement of Microsoft's Surface line of tablets. There are two different models, the "Surface" which is thin, light, and ARM-based, and the "Surface Pro" which runs some sort of Intel Ivy Bridge but that's 33% heavier and 45% thicker.

So since the Surface did not go for the "fast" option, it can be light and have a long battery life. The Surface Pro has the "fast" option, but it's thicker and heavier and will most likely not have a battery life anywhere near as long as the regular Surface. (The Surface Pro has a 42Wh battery, which is typical of an Ultrabook-class device.)

To be clear, I'm not saying that any one of these options is inherently "better" than the others. It's a use case by use case basis. But the reality is that I want one combination of these options for some use cases, and other combinations for other use cases. (Light, long battery, but slower when I'm running around and fast when I'm stationary getting real work done.) So that's why there will also be a market for all three combinations of options—one type of solution will not "replace" the others (or all three).

The only future caveat to this is that maybe (maybe!) we'll get to a point where all three are "good enough" and it won't matter. But clearly we're not there yet. (If we were then Microsoft wouldn't have ARM and Ivy Bridge options—they'd just have the single "good enough" tablet.)

Join the conversation


Send me notifications when other members comment.

Please create a username to comment.


Don't published apps, especially when both the apps and the remoting protocol are touch enabled, break this triangle? Sure, you have to be online, but that may be an acceptable requirement in many scenarios.



Well sure, published apps break the triangle, but that's a special use case. They also break the balance for desktops too. And of course, they introduce their own triangle of user experience, bandwidth, and host side CPU.

Bottom line though, yeah, if you want to use a tablet like a thin client, you can potentially have access to more CPU power on the back end, but that brings its own set of limitations and complexities.


I like it.

I would argue (as eluded to it meaning different things to different people) that more specifically, "fast" is highly relative to the workload.  A Blackberry in 2007 had all 3, but the workload then wasn't large screen/high resolution video.

I think the importance of maintaining relativity to all 3 points of your triangle is key since the tablet/slate form factor is being challenged both up (Ultrabooks) and down (Mega-Smartphones like the Galaxy Note).  

As the Surface illustrates, and just as we've all dealt with from a VDI/RDSH perspective : What is the use-case?  What is the workload?


I think you are missing something Brian, Windows RT changes your triangle a bit. This is because they have the ability to redefine "Fast" in your triangle. Because the number and types of libraries supported are limited on Windows RT they can cheat a little and offer a higher "Fast" peak for a given amount of battery life because there is simply less going on every time the user interacts with the OS.

To me this is one of the most interesting aspects of this Windows All-In strategy, providing a similar speed and usability of interface across multiple types of device, while tuning the types and amount of WORK that happens behind the scenes based on the device capabilities.

I agree that the inclusion of the typeable cover would appear to make this less of a tablet, but the software keyboard is a first class citizen now as well, so the Type and Touch covers become something that you CAN use when you need to, but are invisible to you when you don't.

So far what we have seen is pretty encouraging, I can get by with either Surface device for 100% of my road/home work tasks, and if I need something more I can always use a real PC at home/office. Like you said, a companion, but one that is more flexible than most tablet only devices, and more tablet like than any Tablet-PC or even the Transformer Prime.

That is a big market space, so if they pull off the performance and app catalog story (maybe with some more content announcements to sweeten the pot) this could be a significant player. It will all depend on seeing the same level of attention to detail in software that we see in hardware, but I'm going to reserve my judgement on that until we see more info.