As used today, yes. And that's kind of the point of the original observation. Smartphones are not true supercomputers. But they still smoke a Cray-1, and yet we use them as dumb terminals. What can we use that power for, that filled a room twenty years ago?

Their existing roles have value! But there is a sense of leaving something untapped, when you use them only as dumb terminals. A sense of, what am I missing?

It doesn't make sense to have a supercomputer as your phone. In your phone you value excellent long battery life. Better to offload the supercomputer part to central processing and pipe down a "viewport". Economics always wins.

"Phone" isn't really a term that makes sense for the future. It's just a somewhat anachronistic bundling of microphone/speaker/digital radio/battery (and the tiny bit of logic needed for that to work together).

It does make sense to carry the things that comprise a "smart phone" (as above, +powerful cpu that can scale down to conserve power, gpu, screen/hud/display, camera(s), low-power local-area digital radio (eg: bluetooth), touchscreen (or drawing surface/digitizer), text-input (keyboard), gps, interface for connecting 4k+ display, possibly direct-attached storage), along with a headset, preferably wireless.

I just don't see why this should all be forced into one unit. I'm not sure how much power miracast would require -- but could it really be more than what the led-lit touch-screen uses? Because surely it'd be easier to have big battery+wlan/miracast/bluetooth/cpu stuffed in your pocket, and a small display+touch+bluetooth device on your wrist, with the option to pull out a phablet-sized hi-res touch-screen with it's own battery, but only miracast+bluetooth (maybe a gpu)?

The "smart stick" (the smarphone less the camera/screen/mic) could probably be made to fit more nicely in a pocket if there was no need for a display.

Shape it like a rod, add a powerful vibrator and the thing would even be dual-use ;-) (seriously, though - if it's intended to be kept in a pocket, a vibrator might make sense, as well as a couple of status leds).

The screen is going to be a bit clunky anyway (unless it folds/rolls up) - so no real issue to add a dedicated battery for it.

Maybe I should file a patent for "smart rod".

I've always wanted an on-body cluster ;)

But how come heads-up displays have never come to market in a serious way? I'm not counting Glass, by the way, given the focus on sharing rather than using.

Beats me. I also wonder why OLED screens only seemed to be used for military HUD/goggles -- and not for the civilian market.

I recall one pioneer from MIT bashed glass, pointing to his experience with having a single image in a different focus area/off-center as being uncomfortable for prolonged used (or maybe even harmful).

So that might be one reason (similar to how everyone realized in the 90s that VR made people throw up -- no market).

As for a smart-rod -- I really liked Canonical's push towards a viable use-as-phone-use-as-computer OS. It's an obvious idea that I think is already "proven" by stick-pcs and chromebooks. The cpu of a previous-generation smart-phone is enough for many tasks.

Another "proof" is what's happening with Samsung's Gear VR IMNHO, eg: http://vrjam.challengepost.com/updates/3462-meet-your-mobile...

Not sure if VR might not actually be an area where going wireless might break down though -- Oculus has spent a lot of effort getting the latency down. Still it seems it should be possible to do a lot of the 3d stuff along with head tracking in the head-set, and do more complicated physics modelling on a more powerful device.

Either way, I'm hopeful that we'll see more meaningful separation of devices as bluetooth (as I understand it) keeps getting support for lower power in one end, and higher bandwidth in the other.

Because the truth is a HUD is useless most of the time. What are you gonna put on it? Walking speed? Time to lunch? Random Tweets?

I'd want the same stuff on it as I have on other PCs. Tools or access to tools. And yes, mobile-specific stuff too. Why work in/on wherever using a tablet, when you could instead have a full virtual work environment?

Maybe not walking speed, but perhaps biking speed? Text messages? Caller id (for screening calls)?

Noooooo. Mobile bandwidth is far more constrained that mobile CPU power. Furthermore, radios are the most power hungry components after the screen, another reason to try to avoid them. Data connections are not truly ubiquitous either, so what do you do when you don't have a connection? Lastly, from a data security point of view, if you don't want to hand all of your unencrypted data to a 3rd party, you would have to host the server yourself (in this model the server needs to be able to manipulate your data, which means that it must be able to decrypt any encryption you had on the data connection itself).

So no, I don't see that as ever being a desirable model for mobile computing. Maybe for some specialised tasks, but for general pocket computing, the power is better placed locally, not remotely.

> In your phone you value excellent long battery life.

No. You and I might want a long battery life, but 'the market' has decided that battery life is way down the list, below 'look like an latest iphone'.

The market has standardized on shiny glass brick, as thick as a pancake, and tiny batteries to match. Batteries that are less than 3,000 mAh and that barely last a day, two days, tops.

If the market thought an excellent battery life were important, there would be phones an inch thick, with a 10,000 mAh battery that lasted a week instead of the latest generation featuring 8-core, 64-bit, gigahertz+ processors.

Economics always wins. Unfortunately, economics is not always on your side.

But you already have the performance of a 1980s supercomputer in your phone. It doesn't have to make sense.

But you have a limited power supply.

I can easily carry a power brick. Or a hydrogen fuel cell like they use for power tools.

But most people either can't or don't want to.

For me, they're mostly missing privacy. That is, they're not securable from supposedly trusted providers. So for me, a precondition would be full control of the baseband, and running open-source software. Or maybe for now, using Android kernel to load Debian, and nuking the radio.

I'm sure that lots of cool distributed computing stuff is doable. But I'm not much interested until it's securable.

For those who think I'm crazy, consider what company would accept smartphone-like security for its servers. And OK, upon reflection, some apparently do. But my point stands.

For most people, computation isn't nearly as valuable as communication. Raw computation isn't useful to non-scientists; the smartphone computation power is put there to render games and web pages.

Unless I'm mistaken, a Cray supercomputer generally had a dedicated and consistent power supply, i.e. they're plugged in. Most phones aren't, and battery technology is still not at the point where you could achieve constant performance if you were using the phone all day.

Well one of those games is playing world championship level chess. Pretty cool for something that fits in your pocket.

Not at all. Even IBM Deep Blue chess computer relied heavily on opening and ending chess databases. It's a common thing.

Also does the app rely on a server to do the processing? As many F2P and online games do.