Pocket Wars and Cores

An anonymous reader writes "If I were to ask you what is the most popular processor used in phones and pads, and you said, 'ARM,' you would be correct. Now comes the trick question, 'Who make ARM processors?' Not the ARM Holdings company. They design processors and license their designs to manufacturers. They also have a reputation for creating very low power designs. Interestingly, while almost everyone else was out ramping clocks and power consumption (until they hit a wall), ARM was chugging along addressing the low power end of the market. Now that low-power is all the rage, due to phones and pads, ARM has become quite a bit more popular."

Comments

Too bad!

[Anonymous Coward]

Too bad nobody's making ultra-cheap machines yet.
Why aren't there 50$ SOC systems on the market ? Not tablets , desktops will do, or thin clients.
First post ?

Re:Too bad!

[fuzzyfuzzyfungus]

Your off-brand Chinese importers can hook you up with an ARM-based netbook-esque mini-notebook for $80-$100, depending on exact specs, volume, and the whims of the ebay gods.

Trouble is, in most cases, these will either be running some dubiously-legit(and sometimes questionably well-localized) version of WinCE, or a mildly elderly version of Android. Actual cryptographic lockdowns, in the Apple or Motorola vein, are way outside the budget; but total lack of usable documentation, a confusing proliferation of part numbers, or rampant hardware switching between similar looking models has somewhat retarded the growth of decent sized 3rd-party release groups.

Curiously, the hardware built into these $80-$100, with (lousy) screen, keyboard, and battery doesn't generally seem to show up in $40-$50 versions with DC-in, VGA-out, and USB for peripherals. There are some machines with those specs, like HP's t5325; but the fact that that is a "thin client" and thus "enterprise" instantly doubles the price you'd expect for the specs.

You can also get quite capable hardware in Marvell's *plug line; but those are generally network appliances only, with your only display option being a USB-based Displaylink or similar. That certainly works; but nearly doubles the price and makes for a rather ugly donglefest.

The newer Marvell SoCs do support at least one lane of PCIe, in addition to a raft of other onboard peripherals, so it wouldn't be rocket surgery for an OEM to put out a *plug-esque design with an actual PCIe graphics chip(only a low-end one would really make sense; but even the cheapest PCIe graphics chips available can drive pretty much any monitor that doesn't require dual-link DVI) hanging off that lane. However, that is a bit hardcore to just hack onto an existing *plug board, and, as noted, nobody seems to have done that in commercial quantity.

You can get the cheap-and-nasty "PocketPC of yesteryear shoved into a clone of the EEE701" from any number of mystery OEMs on ebay; but the software will blow and 3rd party firmware support is kind of a gamble.

You can get a *plug-based design, which will have a much peppier ARM core (1.2GHz) and beween 128-512mb of RAM, depending on the exact model, for about the same money(Seagate Dockstars were going crazy cheap for a while, like $10-$20; but that was a firesale of sorts); but those are network-only unless you buy a Displaylink adapter, which pushes you up toward $150-$200, at which point Atom boxes that will run normal x86 OSes with zero hassle and take 1GB+ of RAM start to beckon...

The t5325 is pretty much exactly what you are asking for, except that it is an "enterprise" product, and has a price tag to match. If one could hunt down whatever OEM produces the board inside, and buy 10,000 of the same thing in generic black boxes, those would probably be precisely what you want; but I've never seen any hints on how to do that...

It's a bit more complex than this article...

[rcs1000]

...makes out

There are many, many makers of ARM based 'application processors' and the like: Texas Instruments, Samsung, Apple, ST Microelectronics, nVidia to name but a few. In addition, some people - like Qualcomm with their Snapdragon processor - have licensed the instruction set from ARM, but then have basically built their own core around that.

The nice thing about ARM is that - if you are looking to embed processing power - you can license a core (or two), design them into your own chip and then make it. Said chip can also include a USB controller, or a wireless baseband, or whatever. Intel will not sell you an x86 core for you to design into your own chip; ARM will.

Now: before this thread descends into meaningless ARM versus Intel rivalry, can I point out that the two architectures are optimized for entirely different situations. To say ARM is better than Intel, is like saying a bicycle is better than a ship - it's not a meaningful comparison. If you want to embed processing functionality, or you want low-power (particularly low standby power), then you need ARM. If you need raw processing power, optimised to run desktop or server operating systems, then you'll be wanting x86.

And the reason why x86 is so power hungry? It's because it's on big bits of silicon. And why's it on big bits of silicon? Because it support hyper-threading, out-of-order executon, has hardware virtualisation extensions, has extensive branch prediction, and tonnes of on chip cache.

There is no reason why ARM cannot offer all of these things too (and their Eagle design goes some way to do this). But if you want to do this, then your chip is going to get bigger, and more expensive, and more power hungry.

Over the next five years, we are going to continue to see mobility become more important: and that means more and more ARM cores, and a diminution of the importance of the traditional PC market. ARM has a very bright future - but, I suspect, it will probably have a great deal of trouble getting into the traditional PC space.

Re:I think my first ARM device was a Gameboy Advan

[PhunkySchtuff]

Acorn machines were incredible for their time. Their GUI had concepts that have only been realised in mass market GUIs just recently, the flexibility of their OS and their advanced typographical features were many years ahead of their time. Things like the save dialog for a new file having an icon of the file that you could give a name to and then drag that icon to a folder to save it there (rather than having to navigate to the folder in the dialog). Built in BASIC in ROM (most of the OS in ROM, so boot times were on the order of seconds). I could go on...

No Windows 7 Mobile on ARM

[erroneus]

I read through the article and found it very informative. One thing I didn't realize was that Microsoft will not do Windows 7 mobile on ARM.

That was a surprising statement. I googled on it and found this:

http://www.xbitlabs.com/news/other/display/20090603123741_Microsoft_Windows_7_Will_Not_Support_ARM_Microprocessors.html

This article says "Micrsoft does not believe ARM can deliver the performance needed." To that I wonder "why is everyone else able to make amazing performance happen with ARM???"

Every time I hear another Microsoft shill claim "but this is not Windows, it is entirely new from the ground up" I have to chuckle a little. If that were true, then they wouldn't have any problem getting performance out of low-power hardware if they designed their OS with that in mind "from the ground up." The truth of the matter is that Microsoft simply can't get away from its legacy code and rebuild from scratch. I shouldn't say they can't -- I should say they are unwilling. Apple did it when they went with OSX. A completely new OS and while the transition was painful for users and developers, it was the right choice. I have been saying for nearly a decade that Microsoft should do the same... others have too... but they simply choose not to at every opportunity.

This whole scenario gives me a better understanding of why Windows Mobile isn't catching on even with hard core MS fans. The "desktop experience" doesn't fit in your hand and they simply don't know how to do it any other way.... (Or maybe they are afraid to since MS Bob...)