Is ARM Ever Coming To the Desktop?

First time accepted submitter bingbangboom writes "Where are the ARM powered desktops? I finally see some desktop models however they are relegated to "developer" models with USD200+ price tags (trimslice, etc). Raspberry Pi seems to be the only thing that will be priced correctly, have the right amount of features, and may actually be released. Is the software side holding ARM desktops back? Everyone seems to be foaming at the mouth about anything with a touch interface, even on the Linux side. Or are manufacturers not wanting to bring the 'netbook effect' to their desktop sales? Are ARM powered desktops destined to join the mythical smartbook?"

Comments

Tabtop momentum building

[Anonymous Coward]

Waiting for the release of Tegra 3 Kal-El quad-core equipped Asus Transformer 2 tabtop as a low power general purpose device that I can keep powered on all the time to replace a much more power hungry x86 machine.

Re:Tabtop momentum building

[GameboyRMH]

...then finally the device will become slim enough to have the keyboard built-in without pissing off even the trendiest of Starbucks-dwellers, and we would have come full circle back to the convertible laptop.

Re:Tabtop momentum building

[hairyfeet]

I'm sorry but.....why? WTF would you want ARM on the desktop? Are you living in a mud hut in Zambundi and don't have any electricity to spare for a desktop?

Lets be honest folks, the big selling point of ARM is how cheap it is on batteries. Well guess what you do NOT need when you are inside? Why that would be a battery! See that plug on the wall right in front of you?

Cycle for cycle x86 stomps the living shit out of ARM, it just uses more power to do so than most mobiles can afford due to the fact we haven't had a real breakthrough in battery tech in ages. Well that and the fashionistas at Apple have made iSliver batteries the "in" thing in which means you have to power the thing on a battery the width of a tic tac. I don't care if you put 8 cores on the thing, a bottom o' the line AMD quad, even the low power AMD quads, will stomp the living shit out of ARM. drop in an i series and it isn't even funny how badly it gets stomped.

Like everything else it is about using the right tool for the job. ARM royally kicks ass in mobile, embedded, and in places where you need a device that'll take milspec levels of abuse due to the fact you can run it fanless. X86 kicks ass in desktop and laptop where you want more performance and don't mind giving up some battery life for it. But ARM on the desktop makes about as much sense as stuffing an i series into your phone, that is none at all. The majority of code out there is x86, even on Linux x86 outnumbers ARM code by a pretty wide margin. So unless you just really really REALLY want the Droid version of Angry birds on your desktop it just seems more than a little stupid to be running a mobile chip in a place where you are right beside a plug in.

Re:Tabtop momentum building

[MightyMartian]

The majority of code out there is x86, even on Linux x86 outnumbers ARM code by a pretty wide margin.

This is a bizarre claim, considering the majority of code out there is in C or in higher level languages like Java, Cobol, C# and so on, so technically the processor architecture is irrelevant for most code.

As to Linux, there are small pieces of the kernel written in assembly, but these have been rewritten so Linux can run on a number of non-x86 platforms. The vast bulk of Linux and its userland tools are written in C, so the underlying architecture is irrelevant. Want to run emacs on an ARM variant of Linux, well, just bloody well compile it for that ARM processor.

Re:Tabtop momentum building

[jd142]

I'm sorry but.....why? WTF would you want ARM on the desktop? Are you living in a mud hut in Zambundi and don't have any electricity to spare for a desktop?

Lets be honest folks, the big selling point of ARM is how cheap it is on batteries. Well guess what you do NOT need when you are inside? Why that would be a battery! See that plug on the wall right in front of you?

You know, it's just possible some people might want to conserve electricity. Or even shave a couple of bucks off the old electricity bill. Just because you can use a resource, doesn't mean you should. I have running water, but I don't just leave the faucet on all day in case I might want a glass of water.

I don't know, but if you had one of those little portable solar cells, could you just power an arm laptop anywhere?

Re:Tabtop momentum building

[icebraining]

This is a bizarre claim, considering the majority of code out there is in C or in higher level languages like Java, Cobol, C# and so on, so technically the processor architecture is irrelevant for most code.

That's fine for OSS (and Debian for example has decent support for ARM), but try to convince some publisher of some proprietary software you need to use to port it.

Re:Tabtop momentum building

[wonkavader]

I want silence. COMPLETE silence.

I want a computer you can barely find, it's so small and unobtrusive.

I want a computer so cool it can be covered in papers and crap without me worrying about it overheating.

I want devices that are dirt cheap to buy and dirt cheap to run, because I want them in every room, on all the time.

I want ARM.

Re:Tabtop momentum building

[Nutria]

Or a Lenovo propped up on it's docking station and connected to an external kb and LCD.

Re:Tabtop momentum building

[m50d]

This is a bizarre claim, considering the majority of code out there is in C or in higher level languages like Java, Cobol, C# and so on, so technically the processor architecture is irrelevant for most code.

Speaking as someone who has ported an actually quite well written (relatively) game interpreter from x86 to arm: bwahahaha. For java/C# it's feasible (although even then, four out of five programs will require at least some superficial code changes), but porting anything written in C is going to be a headache at best, and more likely a complete nightmare - and that goes doubly for C++ actually, at least the way it's commonly written.

Re:Tabtop momentum building

[m50d]

If cheap is what you want you'll get it from x86 rather than ARM. Look at some of the fanless VIA systems; they'll draw more power than their ARM equivalents, but they're cheaper, more powerful, and just as cool. (Ok technically if they're drawing more power they'll be warmer, but they're in the same category of "I don't have to worry about how hot it is").

Re:Tabtop momentum building

[catmistake]

I wanted someting like this too, and I was sort of obsessing over different Atom D525 systems... by all accounts fanless, silent Atom systems are now buildable. But then I read how the AMD E-350 just spanks Atom in processing power... and I began to realize I just don't care so much about having the best possible power efficiency... what I care about is cheap powerful systems, and if it sucks less power than a lightbulb or even an unused but plugged in wallwart... then that is neato, but not of primary concern in something that is a relatively permanent fixture in the home, with plug in power available.

Re:Tabtop momentum building

[backslashdot]

The keyboard should flexible-ish and built into the tablet pc's cover.

Re:Tabtop momentum building

[TheSunborn]

And what does this have to do with arm?

An intel atom chip could give you all that*.

*Within a few wats, but thoose 2 or 3 wats are not the reason such a computer don't exists. 

Re:Tabtop momentum building

[hairyfeet]

Then what you want already exists in x86 form, its called Brazos and to use that old John Romero quote it would "Make ARM its bitch" hands down. No fans, small enough you can just bolt it to the back of the monitor, yet it has a dual core APU with Radeon 6310 so it'll stomp at multimedia and HD. Oh and it MAXES OUT at 28w, and that is if you slam the living hell out of it, most of the time its below 6w, which is less than the modem you use to get the net into your house.

So sorry, already exists without having to deal with porting everything to a cell phone chip. I have sold several of the laptop version and its damned nice and gets around 6 hours on a 6 cell battery. i liked it so much I ordered a EEE version for myself, that baby will hold 8Gb of RAM and only cost me $340 counting the extra 4Gb stick. I thought about 8Gb but WTF? When will I need 8Gb in a netbook? Great for multimedia BTW, and has both HDMI out and USB 3. Gotta love the new AMD APUs, sweet, fast, and cheap, just my combination.

Re:Tabtop momentum building

[Paracelcus]

Anybody remember Acorn? They didn't do so well!
Arm on the desktop? been there, done that, went broke.

Re:Tabtop momentum building

[tengwar]

Could you give a bit more detail on the difficulties? This seems a bit surprising.

Re:Tabtop momentum building

[HappyPsycho]

I think you are looking at it from the same perspective I would, as a power user.

My phone (HTC HD2) can easily run everything everyone else my house uses on a daily basis fine (browser, mail client, video player, music, word processor), my usage is different but only my development environment and certain games are what force me to have a beefy PC.

What benefit other than established code base is there for sticking with x86 for the majority of computer users? Its been pretty well established that for most users a PC that's 3 years old (or possibly older) would easily suffice. Smaller form factors, less power usage and less heat are becoming important factors for almost all users seeing as speed isn't noticeably increasing for the vast majority of users.

One possible benefit to power users for going low power is longer UPS battery times. If your UPS can last hours instead of minutes would you care?

Re:Tabtop momentum building

[m50d]

ARM is admittedly easier than many things - you're still 32-bit and usually don't have endianess issues. So the main problems are sizes/alignment of system data structures - things like (network) host information, or even something as simple as date/time vary. Of course you should be using the types opaquely, but it's very easy to accidentally stuff a time_t into an int, and even if you are being careful you can miss implicit assumptions you're making (one of the problems I saw was code not dealing with an array in the network host data structure being of length 0). In this particular case the biggest source of problems was multibyte character handling differences (it was a Japanese game). When porting old code away from x86 you can also get a whole slew of floating-point related bugs, because double on x86 is really x87 extended double (I only encountered one minor problem in this case, it didn't do much floating point, but for other software it would be a bigger issue). Also see the recent glibc memcpy vs memmove issues for something that can come up even on x86-64 vs x86-64.

Once code's been ported to two or three architectures these problems don't tend to come up any more (because the first couple of changes reveal all your implicit assumptions that could be broken), and that's true of a lot of open source projects. But any code that's only ever run on one platform will have portability issues. You don't have to take my word for it - try it yourself, pick a random project that doesn't release non-x86 builds off sourceforge and try and build it for arm.

Re:Tabtop momentum building

[negRo_slim]

28W is still quite a bit of power.

Re:Tabtop momentum building

[Lennie]

Well, there are a few thing you might not have thought of:
  -most people really don't need a fast machine.
- and a lot of people have a device they don't want to plugin in every momeny. Like for example an tablet-device or a laptop they might be using when in the garden.

Re:Tabtop momentum building

[Chas]

You know, it's just possible some people might want to conserve electricity.

Look at the power said system uses to complete any given task.

If you complete a job in 10 minutes and burn 1KW doing it how is that in any way superior to doing the same job in 5 minutes and burning the same 1KW?

Re:Tabtop momentum building

[laffer1]

Those people would probably buy a laptop or netbook anyway. They use less power. The whole point of a desktop is performance. In the old days, some people liked the form factor. Most people only buy desktops because they don't know better, are avid gamers, content creators or programmers.

Anyone who wants a cheap or energy efficient system shouldn't buy a desktop.

Re:Tabtop momentum building

[tengwar]

Interesting - thought we'd got past this stage. Still, I found myself working on some K&R C a couple of nights ago.

Re:Tabtop momentum building

[atlasdropperofworlds]

> Want to run emacs on an ARM variant of Linux, well, just bloody well compile it for that ARM processor.

Just don't compile it on an ARM processor or you'll be waiting forever, or at least until intel releases something that will beat arm down in power usage or when an ARM-based processor comes out that can push 200x more FLOPs.

Re:Tabtop momentum building

[Dogtanian]

Anybody remember Acorn? They didn't do so well! Arm on the desktop? been there, done that, went broke.

Maybe that's a "whoosh" I hear, but... you do realise that ARM *was* originally developed for desktop use and by Acorn themselves at that!

And I suspect that the relative failure of the Archimedes- where ARM first appeared- had more to do with other market factors than ARM itself, which apparently had excellent performance for its time. (Remember also that the Amiga and Atari ST were both popular in Acorn's home market of the UK (and in Western Europe generally) in the late 80s and early 90s, so lack of x86 compatibility wouldn't itself have been the kiss of death).

Acorn themselves ultimately fizzled out in the late 90s- probably victims of the market's standardisation on commodity PC clones by that point- but their ARM spinoff was massively successful.

Re:Tabtop momentum building

[udippel]

My astonishment of the month of September: One mod point only for parent and FP at 232 other comments.
Later on, some ludicrous comments get many more. Maybe the Slashdot community has slowly migrated from the nerds and geeks to the gadgetry crowd??

Okay, now on-topic: Exactly my thoughts. I have been waiting and waiting and waiting for a desktop machine to run ARM. Not a netbook, not a tablet. Oh, yes, I'd also like to have a tablet-based-ARM running Debian; but that's off the topic.
I want to have a desktop as well, stationary, with a keyboard, a 24-inch monitor plugged in, a DVD-drive, space for 2 built-in hard disks, 5.1 sound, a number of USBs to stick in stuff as needed temporarily. For the last 4 years this was a AMD ...-EE with a TDP of 45W. I even paid extra. Now I am developing on a Pandaboard, where that whole board goes with 2.5 W. The Energy-Efficient AMD still needs a fan, and produces quite some warmth. The single bridge - though more efficient than north/south can barely be touched, despite of the build-in heat-sink. And what I am doing is working on text, browsing, e-mailing, P2P, ssh into my servers, playing of media. I had thought the Panda would be sufficient, and it almost is, even. Now I want an ARM, with less than 10 W altogether, in a - if need be - MicroATX casing, with the speakers attached, 6 external USB ports, etc. This should not consume more than 10 W on ARM, while it still does around 40 W on x86.

Re:Tabtop momentum building

[hairyfeet]

Did you miss the if you were slamming it part of the sentence? I've actually worked with these chips and like I said liked them enough I have a EEE netbook with one on the way and i found there was only really two ways to slam it hard enough to get it that high. 1.-Load up a game far more advanced than the machine should be running, like say Starcraft II or Bioshock II with settings maxed, or try to do multi-transforms with Virtualdub on the chip.

Now be honest: How many people are actually gonna try either of those? how many who would seriously think about ARM on the desktop would try either of those? To use a /. car analogy it would be like saying the Prius is a bad car because it lugs when you are trying to drag a boat. Duh don't drag a boat with a Prius!

But if you are talking everyday surfing and basic office tasks, you know the kind of thing one would reasonably use something like that for? We are talking between 6w and 9w, less than the cable modem. It doesn't even really jump when doing HD because the Radeon 6310 takes the hand off and does the decoding in graphics hardware instead of slamming the general purpose CPU.

Anyway here is a nice article about Brazos and gives all the details and some cool benches. TLDR? More powerful than Atom+ION while using less power. For a low power desktop it is just about as nice as one could ask for.

Re:Tabtop momentum building

[smash]

I think the point is this: desktop CPUs are overkill for most people now, ARM is almost more than fast enough also. Combine this with the fact that the traditional "desktop" is on life support (more and more portable machines are being sold vs desktops every year) and I see a very limited market for desktop hardware.

Why spend X on a desktop machine, when you can spend x/2 or x/3 on an ARM based machine that you can also take with you when you leave the house?

So, i don't think we'll ever see "arm on the desktop". The desktop will be phased out, and ARM's home turf, the mobile space will take over.

Re:Tabtop momentum building

[smash]

Why not go grab an acorn A4000 or something? :D

Re:Tabtop momentum building

[The_Dougster]

I was able to compile the Hercules IBM Mainframe Emulator for my OMAP3 Beagleboard, and guess what.. it works.

Re:Tabtop momentum building

[udippel]

Hahaha. Does it have / support DVD? 5.1? How about Debian? Did you read my post?

Re:Tabtop momentum building

[m50d]

That'd be the Hercules IBM Mainframe Emulator that already runs on three different architectures (and five different OSes), yes? As I said, once you've ported to a few different systems the problems mostly stop showing up.

Re:Tabtop momentum building

[Froggie]

Compiling is not FLOP heavy.

Also, emacs compiled relatively rapidly (under an hour) 20 years ago; I'm sure it's not that bad on anything modern given they tend to have 50x the MIPS of back then.

Re:Tabtop momentum building

[Ofloo]

I bought desktop because if a part breaks down it's not that much of a hassle to replace it you should get your facts straight not everyone thinks like you.

Re:Tabtop momentum building

[sourcerror]

I only know a little C/C++, I'm more a Java developer, so please forgive my cluelessness, so ...
Is the problem the C language itself, or the programming culture that C has fostered (including books, tutorials etc. )? Is there a language that offer native performance with better portability (like e.g. D) ? As far as I know even in C# you're subject to the whims of the garbage collector.

Re:Tabtop momentum building

[sourcerror]

You assume that porting application just happens magically without any environmental footprint.

Actually it requires quite a bit of developer work, and those developers need stuff like food, gasoline in their cars, natural gas to heat their home, and electricity to power their computers. And it this might exceed the savings by ARM desktops. (Also this means getting rid of your old x86 machine). If this whole thing takes off, the more likely scenario is that some people will have both ARM desktop and x86 one as well, as they run incompatible programs, just like we have with gaming consoles now.

Re:Tabtop momentum building

[m50d]

That the language exposes native datatypes (and doesn't define them strictly) is a big part of the problem. Probably bigger is the system API though; different systems provide different APIs, and POSIX is nowhere near detailed or extensive enough to let you write programs. Another problem is that the language was never really written with its behaviour fully specified. If you look at the JVM it does an astonishing amount of work to provide a consistent memory model across all platforms. Also you'll know that e.g. the Java filesystem API (at least pre-v7) is quite weak, because that's the only way to get something that can be implemented consistently cross-platform - and even then it's easy to make java filesystem access non-portable.

As for performance, for modern-sized programs a garbage-collected language will probably perform better than native code. .net will let you compile to native code if you want, and if you're really that worried about garbage collection overhead then use a modern functional language like OCaml or Haskell, where referential transparency makes garbage collection easy. (The modern telephone system is basically written in Erlang; real-time performance and garbage collection are not incompatible). If you want a native language in the sense of having native datatypes then by definition that's going to be nonportable.

Re:Tabtop momentum building

[dhasenan]

C's builtin types have architecture-dependent sizes. In Java, you write 'int i;' and get a four-byte integer. In C, your only guarantee about an int's size is that it's not smaller than a short (and likewise, a short is not smaller than a char, and a long is not smaller than an int).

Another common problem for portability is byte order. The arthimetic ways of getting individual bytes from an integer are portable, but if you try reading an integer array one byte at a time, you'll get different results on x86 and PowerPC processors. Most other languages don't let you do this sort of thing.

Re:Tabtop momentum building

[The Snowman]

I only know a little C/C++, I'm more a Java developer, so please forgive my cluelessness, so ...
Is the problem the C language itself, or the programming culture that C has fostered (including books, tutorials etc. )? Is there a language that offer native performance with better portability (like e.g. D) ? As far as I know even in C# you're subject to the whims of the garbage collector.

C++ is not as bad in this regard since it is higher level. However, in C, there are some subtle issues that can come up. While it is lovingly called "portable assembler," it does expose enough low-level details and it can be configured to ignore warnings that involve a gun pointed at the developer's foot. For example, arrays are not pointers, although they are treated as such in some ways: this can cause subtle bugs if one uses pointer tricks on certain arrays. Type conversions may be valid on a common architecture such as x86, but disastrous on ARM or x86-64. Structure packing (in Java, think of it as the memory layout of all the member variables of a class being sent between JBoss and a client, except there is no Spring or serialization... raw bits on the wire only please) can cause misalignments between architectures, as can endian differences (although most libraries should be using network byte order).

The flipside is that it is easy to write portable C, assuming one turns on (and pays attention to) compiler warnings, uses libraries built for multiple platforms, and generally keeps one's head out of one's ass.

Re:Tabtop momentum building

[unixisc]

Honestly, I don't see ARM as a replacement for x86 for either the Windows or Linux markets. However, one place where I think ARM could replace x86 would be in Macs. Currently, Apple supports a whole bunch of apps for both iOS and Lion, which are ultimately different due to each existing on different CPU platforms. If Apple moved OS-X to their A5 or A6 processors, which are now 4 (or 5) core, then that would be a common platform for both Macs and iPads, and it would thereby enable Macs to natively run all the apps that have been developed for iPad (but not Windows 7 or 8). That would enable Apple to push Macs to iPad owners who don't own Macs, and thereby expand their market in a not insignificant way.

So that would be the only case where it makes sense to have an ARM based desktop (conversely, had IBM or Mot/Freescale had plans to have lower power versions of PowerPC when Apple needed it, the current iP*ds would have all been based on PPC, and Apple could have had all that already).

Re:Tabtop momentum building

[couchslug]

Buy and repurpose a bunch of thin clients off Ebay and you can have pretty much that right now. This guy put a bunch of useful info up which can greatly assist selection:

http://www.parkytowers.me.uk/thin/

Re:Tabtop momentum building

[Pentavirate]

You can always try the Guru plug Display. At $199, it's not a bad price.

Nope

[Anonymous Coward]

No, they won't. All the claims of ARMageddon have been bullshit every year it's been claimed (which is quite some time now). Most people will still prefer to get a higher powered x86 system even if they don't always use the power, plus the fact that most legacy software is x86 only and will pretty much never be ported to ARM is another reason that most won't switch.

Re:Nope

[binarylarry]

"Legacy" only exists for shitty proprietary software.

Re:Nope

[V!NCENT]

Lol...

You probably missed the Windows 8 presentation.

Not only does Windows 8 and the upcomming Office run on ARM, there is already a production ready ARM laptop that's going to be sold.

Image larger than iPad battery life and weeks of standby, a full HD resolution, accelerated x264 full HD video playback. Internet Explorer 10 full acceleration and DX11.

No fans. No noice. No overheating on your lap. Dirt cheap. Light. Fast for desktop use.

The ultimate family laptop, for every family member.

Re:Nope

[macpacheco]

>Not only does Windows 8 and the upcomming Office run on ARM, there is already a production ready ARM laptop that's going to be sold.
>Image larger than iPad battery life and weeks of standby, a full HD resolution, accelerated x264 full HD video playback. Internet Explorer 10 full acceleration and DX11.

Right design, wrong operating system. Scratch Windows, add Linux.
I have been running Linux only on x86 for the last 12 years.
In the last few, I even got rid of flash and other proprietary binary only linux software.
I don't want an ARM desktop, I want an ARM laptop (not a tablet, not a netbook).
The only external connections I would like to see that would make my laptop feel like a desktop is many e-sata connectors, so I can hook up many external hard drives when I'm home/in the office, and no, gigabit ethernet isn't a replacement for that.
So I could go 100% Linux/ARM. Specially since we'd expect using an NVidia Kal El CPU, that NVidia would release an ARM version of their proprietary graphics driver, and that's if nouveau isn't almost perfect by then.

Re:Nope

[Anonymous Coward]

You probably missed the Windows 8 presentation.

No, I didn't. Yes, it runs on ARM but no one is going to run an ARM desktop. The ARM verison is for tablets which is where the software legacy doesn't matter as much.

Not only does Windows 8 and the upcomming Office run on ARM, there is already a production ready ARM laptop that's going to be sold.

That would be awesome if Office was the only piece of x86 software widely used. Get back to me when Photoshop, Final Cut Pro, Maya, AutoCad, Visual Studio, AAA games, etc have ARM ports.

Image larger than iPad battery life and weeks of standby, a full HD resolution, accelerated x264 full HD video playback. Internet Explorer 10 full acceleration and DX11.

And a complete dearth of apps that people expect to run on Windows. TOTAL FUCKING WIN!!! Oh wait...

No fans. No noice. No overheating on your lap. Dirt cheap. Light. Fast for desktop use.

No overheating? So that's why the underclocked ARM processor in my Galaxy S can get nearly as hot as my laptop?

The ultimate family laptop, for every family member.

And runs none of the games, etc that most of the family members will want.

Re:Nope

[V!NCENT]

I don't want nVidia anywhere near my Gallium3D accelerated Linux desktop.

Now that the S3 floating point patent got prior-arted, I am going to enjoy some actually working GPU (minus power management, but Bridgman (ATI graphics driver manager guy) already said that internal open source goodness is on the way).

Re:Nope

[MikShapi]

I wrote a long post recently on this.

I think MS going this way will be a much bigger landscape changer.
Consider:
1. ARM is already powerful enough to run modern desktops (quad-core >1GHz parts). Nowhere nearly as zippy as core-based parts, but functionally sufficient.
2. Jan2007, the PHONE was a DEVICE. Dec2007, the PHONE was an APP. This is about to happen to many desktops.
3. MS announced metro will be underlying UI, but that a traditional desktop windows7-like interface can be launched as neccesary.
4.
i. Apple stack: SoC->iDevice[MOSTREVENUE]->OS->App Store,Developer Community,Cloud App Package,Global Carrier Relations

ii. Google stack (inc. Motorola): stack: Baseband->SoC->Android Device->OS->App Store,Developer Community,Cloud App Package[MOSTREVENUE],Global Carrier Relations

iii. MS (inc. Nokia) stack in 2 years: Baseband->SoC->Win8 Device->OS->Cloud App Package,Global Carrier Relations,DESKTOP developer community brought over by unifying mobile/desktop platforms.

Microsoft has a one-up tho (on Google, not on Apple, Apple's doing their version of the same thing): desktop-in-an-app.

We will, of course, judge by execution (microsoft doesn't have a particularly good track record of doing NEW stuff), but it definitely hauls their powerhouse strengths - desktop & Office - straight in the middle where the battle is still raging. Might even make them relevant again.

Re:Nope

[smash]

Thats why the iPad is doing so well i guess.

Re:Nope

[adolf]

Cool!

Hopefully it'll gain enough popularity that it can survive better than than the Itanium, Alpha, PPC, and MIPS ports of Windows.

I'm not holding my breath, though.

Re:Nope

[Anonymous Coward]

And how many run "Final Cut Pro, Maya, AutoCad, Visual Studio, AAA games, etc"? Especially on laptops.

Face it - apps that 98% of people expect to run on windows are browser, email, IM, office, media player and casual games.

And the other 2% can have their x86 desktops.

Look on eBay

[jimicus]

Look on eBay for an Archimedes.

They're rapidly becoming a collector's item, but they were on the desktop in 1987.

Re:Look on eBay

[damburger]

They had a surprisingly modern looking desktop, that booted in seconds. It would be interesting to see where the platform would be today had it taken off in a big way.

Re:Look on eBay

[jimicus]

The OS itself is still around today, after a fashion. But time has not been kind.

IMV, a fast boot cannot compensate for a spectacular lack of features you'd expect to find in a modern OS. It's a single user OS with co-operative multi-tasking rather than pre-emptive, there's no protected memory or swap support, it's single-user.

Re:Look on eBay

[damburger]

I think its safe to assume that, had it been developed as a mainstream OS in the intervening time, it would've gone to pre-emptive multitasking as soon as the hardware permitted it.

Most of my fond memories are of the interface; the consistent and effective use of three mouse buttons, the innovative save dialog and the way in which applications were packaged (which, honestly, I know people thought was invented with Mac OS X)

Re:Look on eBay

[mountaineer76]

Totally agree, just compare RISC OS 2 or 3.1 to the equivalent Windows version in 1988 or so, Acorn was streets ahead. Still have a 200 mhz RISC PC sitting next to my desk, it's a nippy little beast boots in a few secs or so....

Re:Look on eBay

[Anonymous Coward]

I find it funny, yet sad, that people have forgotten that the A in ARM used to stand for Acorn. I was there, in Cambridge, during the time of the first ARM CPU's development. Friends of friends of the people who worked to create it. At the time it was by far the most powerful desktop and it's various OSs (RISC OS 2+.. Arthur was always a stop-gap OS :) ) were far advanced over everything else available at the time.

I still use a RISC PC today. Also, one of the best case designs ever - it's practically infinitely expandable!

Re:Look on eBay

[rolfwind]

Eh, our 386 computer looked about the same, I'd say it was par for the course aesthetically.

Re:Look on eBay

[newcastlejon]

I came here just to post this. It's not that ARM is ever coming to the desktop; it's already been and gone. By the way, am I mistaken in thinking some of the later RISC PCs used Intel ARM processors?

It's also nice to see that someone still thinks back fondly to Archimedes machines as I do.

Re:Look on eBay

[jimicus]

IIRC there was an upgrade based on a DEC StrongARM processor available for the RiscPC. Not sure if there was ever an Xscale upgrade but Castle Technology had a few systems built based on such a chip.

Damn fast, they were. I used them a couple of times at university - I wish I'd known they were binning them, I'd have grabbed one. Came in one day and found the lab had been re-equipped.

Re:Look on eBay

[damburger]

Bollocks. Best case, your 386 was running this: http://en.wikipedia.org/wiki/Windows_2

Re:Look on eBay

[magnusk]

no protected memory

I keep hearing this, but it's not true; RISC OS had protected memory. Try writing to another app's memory from user mode, or writing to VIDC registers from user mode. But some important areas weren't protected, e.g. the ARM vector table.

Re:Look on eBay

[newcastlejon]

Yeah, it was Iyonix using the XScale that I was thinking of. Still, it's nice to see people still talking about them, like BeBoxes. Pity about Phoebe, though.

Re:Look on eBay

[Anonymous Coward]

Windows v2

RISC OS v1

I don't see a whole lot of difference.

Re:Look on eBay

[Anonymous Coward]

Maybe BeOS would be a better comparison. That was also a great OS that booted in less than 10 seconds. In fact, it didn't even have the hourglass cursor present in all of the other operating systems because it never the system unavailable to the user.

Re:Look on eBay

[SimonTheSoundMan]

You're way off. RISC OS 3.6 was around when when you had your 386. RISCPC and A7000 with their nifty 66MHz ARM processors at this time too.

Re:Look on eBay

[sznupi]

Try ROX Desktop, if you still miss RISC OS? Though it also doesn't seem very active; and, being a very mouse-driven environment running on top of *nix, perhaps it was just too awkward of a beast.

I wonder how it is in places which had "very theirs" computer system at some point - how many people (among general population) would go nostalgic when seeing, in the case of UK, some BBC Micro or Archimedes...

Re:Look on eBay

[SimonTheSoundMan]

I have a Phoebe case in my attic.

Acorn were well ahead of their time. They made IPTV set-top boxes and was known as Pace. Psion their sister company made PDAs and even netbooks, "netbook" being their trademark.

Re:Look on eBay

[sylvandb]

"the hardware permitted it" ? What, it didn't even have interrupts? It was no challenge to do pre-emptive multitasking on a 8080 or a Z80 circa 1979. It's hard to believe it couldn't have been done on ARM 8-10 years later.

Re:Look on eBay

[sznupi]

Psion their sister company made...

...also EPOC aka Symbian, w00t!

Re:Look on eBay

[Dogtanian]

Had it been developed as a mainstream OS in the intervening time, it would've gone to pre-emptive multitasking as soon as the hardware permitted it.

You do realise that the Amiga had full, no-nonsense pre-emptive multitasking on its launch in 1985- and that was based around a 68000.

As the other reply said, it had been possible for years. Surely it would have been doable on the ARM/Archimedes?

Re:Look on eBay

[Dogtanian]

IMV, a fast boot cannot compensate for a spectacular lack of features you'd expect to find in a modern OS.

Apparently much of the OS was stored in ROM, which would explain the fast boot, though it has its disadvantages.

Re:Look on eBay

[Anonymous Coward]

I think its safe to assume that, had it been developed as a mainstream OS in the intervening time, it would've gone to pre-emptive multitasking as soon as the hardware permitted it.

What hardware limitation prevented pre-emptive multitasking? All you need is a way to interrupt the CPU, and all CPUs support interrupts. Amiga OS has pre-emptive multitasking in 1985.

Re:Look on eBay

[jimicus]

Much of the OS? Pretty much all the important bits - an Archimedes would boot quite happily with no hard disk or floppy drive (but it wouldn't be terribly interesting).

There were modules that were stored on disk and could be dynamically loaded and unloaded, much like you get on Linux today. The information on Wikipedia has been somewhat idealised from how it wound up working in real life; these modules frequently provided extra functionality and occasionally you'd find applications that depended on a particular version but didn't provide the actual module so you'd have to hunt down not just the module but the right version of it elsewhere. Certainly up until RISC OS 3.1, the OS didn't have much in the way of a centralised library of modules, nor did it have any means to dynamically load them on demand. The application had to do that. My solution was simply to edit the boot sequence so all the modules that I was likely to want were loaded at boot.

Sort of like the DLL hell you had in the days of Windows '9x, but you couldn't really put the system in such a state that it couldn't easily be rebooted. Usually the worst that would happen is you'd load the wrong version of a module and an application would fail to run.

Apparently later systems were built with the OS on flash ROM, allowing for updates without taking the computer to bits.

Re:Look on eBay

[damburger]

There is doing it, and doing it well...

Re:Look on eBay

[Archibald Buttle]

The assertion that RISC OS would have gone to pre-emptive multitasking as soon as the hardware permitted it simply isn't backed up by the facts.

Acorn went through many generations of ARM processors in their computers, from the ARM 2 through to the StrongARM. Acorn themselves had two different operating systems running on ARM chips that featured pre-emptive multitasking and memory protection in the form of ARX (which got canned before the first Archimedes was launched) and RISCiX (their Unix variant that was sold commercially). The hardware permitted it, Acorn just failed to deliver such features in their mainstream OS.

In the final days of Acorn they announced yet another OS that would have featured pre-emptive multi-tasking, memory protection, and QoS features, to be called Galileo. Whilst it was not made clear, and no technical details of any significance were ever released, this seemed to be a different and separate OS development to RISC OS, and likely would not have been compatible.

Re:Look on eBay

[david.given]

'Not kind' is an understatement. The RISC OS kernel is a really unpleasant pile of kludges.

A few years back I wrote an experimental RISC OS kernel workalike (obligatory link). In the process I found myself delving way too deeply into the core of how RISC OS worked and it's really not pretty. It's got such misfeatures as: a memory allocator that allows you to allocate small blocks of memory from inside interrupt handlers --- it does this by tracing the stack to try and figure out whether it's been called reentrantly; hardware-specific hacks built into what should be generic operating system modules (FileSwitch runs programs by fiddling with the supervisor-mode registers); no real device driver model; about half a microkernel to address the fact that there's no real driver model; global variables everywhere (pretty much everything expects to read a byte at 0x00000108 in order to test interrupt status); APIs which pass flag bits in the upper two bits of addresses; and hysterical raisins everywhere, such as having multiple concurrent system heaps. Plus of course, the whole thing is an unreadable mess of uncommented ARM machine code.

That said, it does all work, and some dedicated lunatics have RISC OS running on the BeagleBoard, so a Raspberry Pi port shouldn't be too hard. (Bear in mind that RISC OS is not open source, though; it's got a look-but-don't-touch shared source license.)

What most people think of when they talk about RISC OS is the GUI, which is a work of genius, although now very dated. (There's no keyboard accessibility, for example. Without a mouse you don't use it. This is something Microsoft got right from the very first, and the early versions of Windows were a joy to use with the keyboard.) I wish that modern systems were half as consistent and elegant as the RISC OS GUI. There have been a few attempts to recreate it with modern systems --- ROX Desktop, for example --- but they never got anywhere, alas...

Re:Look on eBay

[Anonymous Coward]

Bullshit. RISC OS 3.6 didn't come out until 1992, the same year as the Pentium.

Re:Look on eBay

[AmiMoJo]

Amiga OS had pre-emptive multitasking in 1985 on the desktop, and incredibly the latest version does offer some memory protection.

One thing I did like when I was a kid was that the RISC OS BASIC interpreter was damn fast.

Re:Look on eBay

[sylvandb]

What is "doing it well" supposed to mean?

When you have a hardware interrupt and a stack to save registers and a stack pointer to switch to a client stack then preemptive multitasking works great. There is nothing "not well" about it and to prove it we have multiple decades of systems doing preemptive multitasking very well with even less capability than a Z80. Just because Apple and Microsoft (and evidently RISCOS) avoided it does not mean it could not have been done well.

Oh, and don't bring up an MMU and/or memory protection or even priviledged instructions. Those have nothing to do with doing premptive multitasking "well" even if many people do confuse them.

Re:Look on eBay

[JasterBobaMereel]

I remember the PC salesman coming to the college I was at, with the cutting edge latest Windows PC, which they ran Excel on with a load of Whizzy macros to show the speed ... and they ran the same one on a new(ish) Mac, at much the same speed, then the Maths Department head came down with his rather old Archimedes and ran up a PC emulator, with Excel and loaded the same macro .... and wiped the floor with both of them ...

A machine ahead of it's time, but did not run Windows and was not a Mac and so it died ....

Re:Look on eBay

[jimicus]

I suspect your Head of Maths was playing a trick.

The reason I think this is that the PC Emulator on the Archimedes emulated something broadly equivalent to a 186 with 640K RAM. Later versions supported EGA graphics. It was useable for fairly undemanding DOS-based applications but I don't think it could run an Excel macro because the first version of Excel to support macros was version 5 - that came out in 1993 and would have required Windows.

A reasonably cutting edge PC of the time would have been a slow Pentium or (more likely) a fast 486 with maybe 4MB RAM. I really don't think that would have been physically possible to emulate in software on any Archimedes.

However, what is possible is Aleph One produced a co-processor board for the Archimedes that had a 486 chip fitted. Plug that in, you'd fire up RISC OS, run a small program and abracadabra! You could run Windows quite happily. Were I a mildly mischievous head of Maths who'd just upgraded his moderately elderly Archimedes with such a card, I would be very tempted to bring it down and wipe the smile off that salesman's face.

Re:Look on eBay

[JasterBobaMereel]

..You are probably right ... but that means a 486 card in Archimedes still outperformed a dedicated PC (which was a fast 486)...?!

Re:Look on eBay

[jimicus]

That "state of the art" PC - it wasn't RM, by any chance?

Why?

[dukeblue219]

Seriously, what is the reason for having a desktop ARM computer? Power consumption? I don't think there's a very large market for people who will settle for tablet-like performance in order to save a few dollars a month at most on electricity compared to existing low power processors. People with power grid problems will want something that runs on a battery anyway, and a tablet/netbook makes more sense there.

Is it just for something fun to play with? Something small and portable? You can always get a small ARM tablet and hook up the HDMI to a monitor if it's the full size display and keyboard you're missing.

Not sure what touch interface has to do with anything. That could be just as easily implemented with any architecture, and it's maybe the ONE thing I agree with Steve Jobs about -- touch does NOT work as a viable input method for a desktop.

Re:Why?

[hedwards]

That was my though, I have a Zacate processor in my Laptop which would be far more suitable for a desktop than your average ARM processor. ARM would ultimately face the same uphill struggle for acceptance that Intel's MERCED did when AMD whomped them with their AMD64 architecture. Changing instruction sets isn't easy to do and the main reason that folks tolerate it with mobile processors is that it saves them so much scarce battery life. Switching to one of AMD's mobile offerings would pretty much eliminate that concern. My laptop for instance uses 25w with everything maxed out. You're not likely to save enough power going with ARM on a desktop to make it worthwhile.

Re:Why?

[rve]

I have an ARM based laptop. It's fanless, in fact, it has no moving parts at all other than the hinge of the screen, and goes for a day or two of regular use between recharges. I'm not convinced "the desktop" has much of a long term future at this point... i think it will go the way of the workstation.

Re:Why?

[SendBot]

... it's maybe the ONE thing I agree with Steve Jobs about -- touch does NOT work as a viable input method for a desktop.

He may have said that at some point, but you should know by now that Apple changes the kool-aid they serve every so often. He's even spoken at length about merging iOS concepts into the desktop OSX.

I've copied some text straight from the apple web site for the "Magic Trackpad" that make touch sound like you're no longer cool without it:

"The new Magic Trackpad is the first Multi-Touch trackpad designed to work with your Mac desktop computer."
"And it supports a full set of gestures, giving you a whole new way to control and interact with what’s on your screen."
"Magic Trackpad gives you a whole new way to control what’s on your Mac desktop computer. When you perform gestures, you actually interact with what’s on your screen. You feel closer to your content, and moving around feels completely natural."

Re:Why?

[lkcl]

well, fortunately there's soon going to be things like the NuSmart 2816, which will have the best of both worlds: Dual-Core 1.6 to 2ghz, 4gb of ECC DDR3 1033mhz RAM... and only about 4 watts for a system (at the 1.6ghz speed).

i'm working towards getting these - and other such beefy low-power CPUs - plugged in to the EOMA initiative:
http://www.openhardwaresummit.org/forum/viewtopic.php?f=5&t=502
http://elinux.org/Embedded_Open_Modular_Architecture/PCMCIA

Re:Why?

[Anonymous Coward]

The OP, and Jobs, is referring to touch screens, not touch pads.

Re:Why?

[SerpentMage]

Yes and everybody will run under powered desktops in the quest to not have to hear the fan...

I run a desktop because I need the power. If it has an ARM in it, so be it. BUT I need the power. When I develop I am going to use multiple screens. When I run my trading software I need a desktop with multiple screens.

What people need to understand is that there is no solution fits all. Some people don't need a desktop, others do. Some people don't need a tablet, others do. Let's all get this straight we will have more choice, not less choice.

Re:Why?

[Bert64]

A few dollars a month for a desktop...
A few thousand dollars a month for an office full of desktops?

The average office worker doesn't do a lot with their computer, and has been doing much the same thing for years... The only thing stopping them from using 10 year old hardware is modern bloated software which is intentionally incompatible with older versions.

There's no reason that the average user's needs couldn't be fulfilled by a low power machine with equivalent processing power to a system from 10+ years ago, with power hungry x86 systems being relegated to the small niche of power users and certain classes of server.

(in short, watch what x86 did to Sparc/MIPS/Alpha/Power, attacked from below)

Re:Why?

[nurb432]

You can always get a small ARM tablet and hook up the HDMI to a monitor if it's the full size display and keyboard you're missing.

That is my thought ( and what I'm doing ), so really ARM is already ( back ) on the desktop.

One advantage to having it on your desktop is that you could have the same 'stuff' in your hand ( phone ) and desktop.

Re:Why?

[Gordonjcp]

Dual-Core 1.6 to 2ghz, 4gb of ECC DDR3 1033mhz RAM... and only about 4 watts for a system (at the 1.6ghz speed).

That's going to be slower than an equivalent x86-based machine, though.

The thing is, it will be slower *but draw one tenth the power consumption*. I want ten of these on a board, with about eight times the processing power for the same power draw as an x86 solution.

Re:Why?

[Bert64]

Exactly, no solution fits all... Your needs are specialised, so you will occupy a niche of people who will continue to buy highend workstations...

For the vast majority of people computers became powerful enough for their requirements many years ago (aside from increasingly bloated software trying to mask that fact), and they are concerned about price, running cost (ie power usage), noise and that the machine is not an eyesore, and even more so are the companies who buy hundreds of desktops for their employees and don't want to buy a noisy, expensive, large and power hungry workstation for someone who's sole business use for it is to write letters.

Re:Why?

[lkcl]

Zacate is 18 watts! that means you have to have a heatsink or heatpipe, fan and other moving parts, as well as much larger power components. by contrast, with something like the NuSmart 2816 if you run it at 1.6ghz then you can get away with 4 watts and that's *including* the ECC 1066mhz DDR3 RAM. a voltage regulator for a stable 4 watt power supply is approximately a $0.50 cents part. it's a whole different ballgame. so the EOMA Initiative, we've set a 5 watt absolute maximum limit, and are sticking to it. http://elinux.org/Embedded_Open_Modular_Architecture/PCMCIA sadly, not even intel's latest 1.2ghz 45nm CPU, the one that's designed for Meego, is suitable, because the CPU's 2.5 watts, the Northbridge IC is 2 watts, whoops there's not enough room to run the RAM ICs.

even the Z510 shows that this northbridge-southbridge strategy is unacceptable. you *have* to go "totally integrated", in order to reach the required power target. once Intel and AMD start doing that, then and only then will they produce a winner.

Re:Why?

[realityimpaired]

There's no reason a Cortex A7 dual core @ 800MHz wouldn't be able to handle both of the tasks you listed with ease. It could even handle basic gaming if you have a discreet video card to handle the load. Most people don't do the kind of number crunching that a modern high end desktop CPU would allow.

The gamer crowd, absolutely. I can fully understand why they would want a high end processor. Even games that aren't that graphics intensive, like Civilization, are very heavy on number crunching. The office crowd, however, could easily be serviced by a low end low power ARM CPU. I could easily replace my desktop with an ARM-powered nettop without adjusting my computing habits at all, and I'm already running a multi-head setup.

Re:Why?

[durrr]

The Zacate is availible for desktop setups. Motherboard with zacate integrated goes for $100-$120, with RAM, PSU, HD and a shoebox for chassi you get a decent windows computer for $200.

That however is quite a bit in excess of the $35 the Raspberry Pi is supposed to sell for. At that pricepoint you can almost start putting them everywhere before knowing why you're putting them everywhere.

Re:Why?

[gl4ss]

the screen is going to consumer a lot of power anyways.

nothing is stopping from running offices with 10 year old sw too. office '97 is enough to run a smb business, proven in practice many times over.

Re:Why?

[Anonymous Coward]

"Seriously, what is the reason for having a desktop ARM computer? Power consumption?"

No, the reason for wanting ARM on the desktop is efficiency: performance per unit of power consumption.
Why would you not want desktop computing power at 1/10 of the power consumption?
It surely would solve a whole lot of the thermal problems that today's desktop PC technology is facing.

The issue seems to be that while ARM is inherently more efficient than x86, the ARM world has been focusing on low power applications, not on the highest possible performance. But there's no principal reason why a (multi-core) ARM cpu that runs at 4GHZ can not be made, and no reason why that cannot be mated with high-speed RAM and a high-speed system bus.

Re:Why?

[__aagbwg300]

I can't speak for the author, but I think he meant "touch screen" and not just "touch." Steve Jobs (among others) have criticized touch screen desktops because of the "gorilla arm" problem. FWIW, I use a magic trackpad at work and it's good for gesturing - especially if you are already used to it, coming off a laptop or tablet.

Re:Why?

[Anonymous Coward]

It implies that the desktop is going away. Development goes a little bit in reverse and instead we have just tablets* that wirelessly connect to our audio equipmentry and HD TVs effectively doubling as monitors in our living room or in the study.

*of course, keyboard is still a must for writing longer passages of text

Re:Why?

[dunkelfalke]

There is actually no need to. I've got a desktop computer with Core2Duo E8400 CPU and a Radeon HD5770 graphics card and the only fan of that system sits in the PSU (and it is almost silent due to its low RPM). The system is quite fast and the power consumption is moderate.
It is absolutely possible to build a silent and powerful desktop system.

Re:Why?

[hedwards]

Right, and the cost you pay is having to recompile and possibly rewrite every application that you want to use.

As I pointed out, Intel thought that they could introduce a 64bit architecture that lacked support for 32bit applications and ended up being taken over AMD's lap for a spanking.

As for thermal dissipation, 18 watts isn't that much, with the right heat sink you can dissipate most CPUs using something like this: http://www.nofencomputer.com/eng/products/CR-100A.php
And ultimately, that's just at peak use, most of the time, the consumption is significantly lower than that.

Re:Why?

[Waffle Iron]

A few dollars a month for a desktop...
A few thousand dollars a month for an office full of desktops?

We have an "underpowered" Intel Atom-based mini desktop in the kitchen as a terminal, which is silent and uses negligible power.

While it would be usable for an office worker, it *is* noticeably slower than my standard Athlon computers. Just about every real-world web page or application launch takes several seconds longer than the normal PCs. Over the course of a month, all these little delays would almost certainly add up to more lost productivity than the additional power cost, especially if the normal PCs had an effective sleep mode that actually gets used when the computer is idle.

(For example, if the average user does 100 things per day that take an extra 3 seconds each, that's 5 minutes per day, or more than an hour lost time over a typical month. Even with low-wage employees, with overhead that lost time could cost the employer well over $20/month, enough to buy power for a couple of standard PCs going full tilt 24/7.)

Of course, you could try to convince every website operator in the world to recode their sites to not be full of bloated Javascript (starting with the site you're reading this on), but good luck with that.

Re:Why?

[bcrowell]

A few dollars a month for a desktop...A few thousand dollars a month for an office full of desktops? The average office worker doesn't do a lot with their computer, and has been doing much the same thing for years... The only thing stopping them from using 10 year old hardware is modern bloated software which is intentionally incompatible with older versions.

In principle I'm in sympathy with you, but in reality there are a lot of problems with your argument.

• You're suggesting using 10-year-old x86 hardware in a medium-sized business environment. This is different from ARM hardware, which is what this discussion was originally about.
• In this type of environment, the total cost of ownership probably consists of something like 50% support, 25% software licensing, 20% hardware, and 5% electricity. If you use ARM-based machines, the first thing I can guarantee you is that some people are going to complain that there's some piece of software they need in order to do their job, and it's not available on ARM. That means you're probably going to need a mixed x86/ARM hardware inventory. That's going to be massively more complex to support than pure x86.
• Suppose instead that you keep a homogeneous x86 inventory, but you keep using machines as old as 10 years. Statistically, your business is probably running Windows. With that hardware mix, you're going to be forced to support lots of different versions of Windows, Office, etc. Again, this makes support more complex and expensive. Since support is the biggest chunk of your TCO, this isn't a good business decision.
• When you use 10 year old hardware, you get all kinds of other issues coming in. E.g., a machine that old may not have a CD drive.

The truth is that hardware is cheap, and workers are expensive. It doesn't make sense to make your workers even 5% less productive in order to save some tiny amount of money on electricity.

What would really make sense these days for a medium to large business would be to stop paying $2000 for every machine and start supplying 80% of their users with new x86 machines in the $500 price range, on a 4-year replacement cycle. What I've observed where I work, however, is that this is difficult to do, for a variety of reasons. Workers who haven't had a hardware upgrade in 10 years feel like when their time comes to finally get an upgrade, this is their one big chance, and they're going to be stuck with their new machine for 10 years into the future -- so they argue for higher-end hardware. IT wants standardization of hardware to make their jobs easier, and since 20% of users do need higher-end hardware, you can't standardize on the low end. Psychologically, IT wants to work with shiny new toys.

Re:Why?

[rve]

You know, in the late 90's Sun, DEC, SGI and yes, Apple, were getting great profit margins on fantastically powerful graphical work stations. Some models sold for $100k or even much more. Swith to an under powered Linux PC just tosave $100k? Not me, I need multiple processor cores, 1600x1200 pixel 256 color displays, 100kbit/s ethernet.
Of course, 5 years later, all of these companies had either abandoned the graphical workstaton market or were struggling to survive. Today, a $800 laptop exceeds the specs above.

I believe 10 or 15 years from now, it will be somewhat unusual to have what you would consider a full powered desktop PC at home or at the office, and so will doing evelopment work for these no longer very common devices be.

Re:Why?

[obarthelemy]

The question could be put backwards: what's the reason for having an x86 computer ?
In my case, I don't game much anymore, I use OpenOffice... apart from dual-screen support, the Pi-B at $35 does everything I need, though more RAM would be nice. Actually getting 2 Pis for fake dual screen will turn out cheaper than my current nettop.
Once iOS and Android get their "big screen" interfaces right (and Android seems well on its way with 4.x), we can even take advantage of all the mobile apps out there, which are typically much cheaper and more accessible than their desktop versions.

Re:Why?

[obarthelemy]

I bought a passive E-350 board from Asus, and put it in a passive, well ventilated case vertically mounted behind my monitor. Temp was 60-65C, so I added a small and whiny fan.

Re:Why?

[Anonymous Coward]

Not the GP, but developing involves more than just typing code into an editor. Compiling, simulation, matlab-style software bring my "workstation" to its knees, and a dual-core 800MHz cpu is not going to cut it. Granted most people don't need that kind of power, but the thing is that they *WANT* that kind of power. Just like how most people don't need a 300hp car, but they want it.

Re:Why?

[Patch86]

I guess it depends if we're including "full sized laptop" in our definition of "desktop" (which pains me- but it seems to be pretty common where the two categories are "desktop" and "mobile" (tablets and smartphones)).

If you include laptops, power consumption is important; it's either longer battery life or smaller, lighter batteries. That's why an ARM-powered iPad has a 10 hour battery life and is still slim and portable. I don't know about you, but I'd happily go for some of that with my office laptop too.

Re:Why?

[fruitbane]

Jobs and Co. at Apple DID introduce a touch-pad mouse, y'know... That's not too far off from a touch screen.

Re:Why?

[CastrTroy]

I mean, if you can run Quake III on Raspberry Pi, I see no reason why you can't solve most people's computing needs with an 1 GHz or less ARM processor. Now, you can't do video editing, but that's not what most people do with their computers. They type email, visit facebook, type up a document, look at some pictures, or watch a movie. These computers are perfect for those uses.

Re:Why?

[Nutria]

Right, and the cost you pay is having to recompile and possibly rewrite every application that you want to use.

I guess you didn't know that Debian is a multi-CPU distro and that they aren't Gentoo?

As I pointed out, Intel thought that they could introduce a 64bit architecture that lacked support for 32bit applications and ended up being taken over AMD's lap for a spanking.

This is true only for the Windows world. (Which is admittedly 90% of the market...)

Re:Why?

[CODiNE]

The distinction is whether or not it's direct interaction on the screen or by using the touch pads. I don't think Steve Jobs nor Apple have ever changed their minds on a touch screen desktop, the Magic touchpad or whatever is enabling their laptop experience on the desktops. Using a multi-touch Macbook with Lion is pretty nice with all the 2, 3, 4 and 5 finger gestures you can play around with. It's all optional and works well with the usual mouse paradigm.

This really seems to be how they're "merging" the OS'. Not the physical interaction parts being merged, but the underlying code and APIs. Apple has been pretty good at understanding that different screen sizes demand different layouts and use-cases, it would be pretty strange for them to make a 27" touch screen given all their talk of gorilla arm.

But then again Apple has a history of trashing stuff they later adopted.

Re:Why?

[Rockoon]

Perhaps you are unaware...

The E-350 is rated for up to 90C.

Re:Why?

[Bert64]

Screens consume a lot less power these days too, a modern display uses a lot less than an old CRT.

Office 97 is no longer supported by its sole supplier...
While it can still do its intended job perfectly adequately, it has a number of security holes that will never be fixed, and it cannot open files created by new versions.

If you intend to connect to a network or receive documents from third parties than office 97 is unusable now.

Re:Why?

[Anonymous Coward]

Jobs was referring to touching the screen as not working because you can't hold your arms out all day. He never condemned trackpads; Apple's been shipping them for decades.

Re:Why?

[Anonymous Coward]

When he was talking about touch not working on the desktop, he was specifically referring to touching the display, not trackpads or trackpad gestures. You know that, right? So why would you post this? Apple has shipped laptops with trackpads for many, many, many years, and given how much R&D they've put into improving trackpad performance and integrating trackpad gestures, etc. with the OS, it kind of makes sense that they would release a trackpad for desktop use. How is this equivalent to going with a touchscreen (which is what the poster you were responding to, and Jobs, was talking about)?

Re:Why?

[catmistake]

Perhaps you are unaware... The E-350 is rated for up to 90C.

hmm... that's good news for me, just made an E-350 purchase... but does that mean an E-350 putting out 90C will last as long as one at 60C or cooler? I guess what I am asking, is there no reason to cool the E-350 below 90C ?

Re:Why?

[Anonymous Coward]

A lot of people on Youtube do want to be able to edit videos though. Then you have all of the musicians editing music, artists editing images, programmers compiling code and gamers playing games. With all of those people combined, that is a pretty significant portion of PC users who need that power.

Re:Why?

[Howitzer86]

One thing I like about x86 PCs is the fact that if you start out with one and you have zero knowledge, you could eventually learn how to do anything you want with it. That's how I started as a teen. I learned everything I could about these machines, to the point that I make money with them. I learned DOS, I learned the internal workings of Windows, I learned how to fix them. And then, as my machine grew ever more powerful and software ever more available online, I learned how to program, I learned how to use 3DSMax, I learned how to use Blender, I know how to setup FL Studio with my Midi Controller, and so on.

Recently I bought Piano for Dummies, which is a software package meant to teach you how to play the piano. It works with my M-Audio keyboard and eventually I'll know how to create music (hint, I love electronica).

I've seen the resources software like FL Studio and 3DSMax requires, something like that isn't going to run on a tablet without being limited in some way. There is 3D software, music software, and painting software on ARM tablets, but they're all toys. You can't make money with them. You can't develop software for them without a PC. There's no challenge and no need to learn their inner workings. Tablets and the future ARM toy computers are a dead end for people.

Granted many folks aren't going to do what I did with mine, but the potential was unknown to me as well. In the pursuit of increasingly dumbed down and locked down equipment, they won't know what they're missing. And their kids especially won't know.

Re:Why?

[Anonymous Coward]

Just had to get some Apple bashing in eh? Did you masturbate while you type that or will you save that for later, when you are crying alone in your apartment.

Re:Why?

[Howitzer86]

How would you know what they do on their computers? My mom is a fairly average computer user, but she has a consumer level video camera and consumer level video editing software. She's often the one filming our family events (birthday's, marriages, etc), but she's pretty much the definition of an average PC user. Soon she won't even need that camera, and will be able to film HD video on tomorrow's consumer level smart phones. Things everyone will have eventually.

Not all computer consumers are simple Facebook and media consumers. People have hobbies and interests, and often that will mesh well with a decent PC.

What I predict is something that's happened before. Thinking that they know what consumers want, companies will push these ARM PC replacements and realize that they won't sell. They didn't sell in the 90s and they won't sell today. Granted, tablets have their place and people love them, but I haven't heard of anyone throwing away a good PC over it. They certainly wouldn't replace their perfectly good x86 PC with a computer running 'tablet' hardware and OSes.

Re:Why?

[Anonymous Coward]

This is why VDI technologies like VMware View are relevant.

1) Put extremely low power devices on desks (thin clients or zero clients)
2) Move the Win7 desktop to the server room/data center
3) Power down the virtual desktops when no longer in use (off business hours ~7pm-5am or so)
4) Power down the vSphere host using DPM during the same time frame

All the while their desktop is being backed by very powerful gear, 24+core servers boasting memory you can easily describe in %s of TB and very likely 10Gb to the servers they are trying to reach. All for 1Mb/s to the user's desk.

Re:Why?

[Dahamma]

If your company has thousands of employees using desktops the overhead of "a few throusand dollars a month" will be pretty insignificant compared to so many other inefficiencies or costs. They could get 10x the savings by charging for coffee, switching to single ply toilet paper, or firing one incompetent employee...

Re:Why?

[Aggrajag]

Windows 8 and Office 201X ARM and this would be a perfect low-powered office setup.

Re:Why?

[lkcl]

actually if you look up that video showing a web browser of an ARM system vs an Intel Atom, it's only towards the end of the video that they show the specs. 1.6ghz Intel Atom, vs Dual-Core FIVE HUNDRED mhz Cortex A9. the performance was about the same.

this is in part due to the x86 instruction set. it's "compact" i.e. it takes up less memory. that was fine and dandy and necessary 30 years ago, but now, memory is not the bottleneck. so x86 @ 1.6ghz is roughly equivalent to ARM @ 800mhz. yes there is the issue that ARM goes for higher latency (less power) whereas Intel goes for lower latency (more power), wider buses etc.

but the NuSmart 2816 has two versions - one is a 32-bit-wide bus, the other is 64-bit-wide... same width as the memory bus for the Intel Atom (64-bit). i think you'll find, therefore, that the 2ghz NuSmart pisses all over some fairly decent x86 systems. we'll just have to see, though :)

Re:Why?

[HappyPsycho]

This is true only for the Windows world. (Which is admittedly 90% of the market...)

And as Windows is going ARM as well, the difference could very well become non-existant.

Re:Why?

[jon3k]

Well, pretty soon we'll have 2.5Ghz (up to octo-core) ARM CPUs so it might be a good time to get your feet wet in the ARM world.

Re:Why?

[Anonymous Coward]

Seriously, what is the reason for having a desktop ARM computer? Power consumption?

Fanless, quiet, powerful enough for many tasks, I have uses for that. And yachts for example have limited power available for a pc, there are shops that specialise in low power consumption computers specifically for that market.

Re:Why?

[Nutria]

But MSFT has already (for good and obvious technical reasons) said, they won't implement CPU emulation to allow existing software to execute.

Re:Why?

[LordLimecat]

Offices (or their tenants) dont usually pay electricity-- not that Ive ever heard of. Possibly its a concern in the server enviornment, but if thats the case Intel makes some new Xeons with 20w power draw (Xeon E3 1220L) which would kick the crap out of those ARM processors.

As for desktop prices, will it run Windows well? What about Office 2010? If not, are you factoring the costs of switching your software suite over to something totally different?

Re:Why?

[HappyPsycho]

Fair point, but if the windows world supports both there is at least in my opinion a strong desire for any proprietary vendors to follow suit.

Re:Why?

[Anonymous Coward]

what is the reason for having a desktop ARM computer? Power consumption?

Yes. Because then you don't need a CPU fan. And because you need so little power, you also don't need a power supply fan. Which then makes the entire computer silent. And then because of the combination of silence and low power needs, you can leave it running around the clock, even in a bedroom (mmm, silence); no more waiting for boot (and the loading time after reaching the desktop), and the power draw when you're not using it is probably still less than many of your other appliances are leeching while "off".

There are further benefits. For example, getting a UPS for it could be very small and very cheap. In fact, it could easily come built in and take a few AA NiMH rechargeables. (They're surprisingly good and cheap these days. Four of them would run your 1-watt ARM computer for several hours, and they're good for like 1500 charge cycles. It won't power your monitor at the same time, but it wouldn't be technologically difficult to have the PSU+UPS tell the OS to throw the system into ultra low power suspend if main power goes off for more than a few minutes - in which case they'd probably last you several days).

All-in-one models could, of course, be extremely small. Look at smartphones and plug computers, for example. Add a hard drive and optical drive, and the device is still not much bigger than a stack of four CD jewel cases. (Or don't add those, and it's only the size of a phone).

And there's no reason you couldn't do a full strength gaming system based on ARM. It wouldn't be teeny tiny, and it'd need a more "normal" PSU to power it. But you could essentially have one of those 150 watt monster GPUs. (It might even be possible for hardware makers to build the ARM systems such that the GPU is only powered on when gaming). This might actually eventually be ideal, to be honest; by replacing the 65-125 watt x86 with a 1-5 watt (probably quad core ARM Cortex A9), you just freed up an extra 60+ watts in your power budget for a beefier GPU. The top end ARM chips coming out in the next six months are comparable to what I was running as a brand new desktop chip five years ago. And the games generally weren't CPU limited then or now, they're GPU limited.

In fact, since Nvidia makes both a line of high end GPUs and a line of high end ARMs, I would be completely unsurprised if such a system existed in the next 18 months. Even if only as a limited run proof of concept for them to show off their tech. (There probably IS a market for "cheap" raw compute power in the form of a few giant GPUs in a small box with an ARM to run linux. That saves a couple hundred bucks in CPU cost and a hundred watts of electricity cost and some cooling cost and so on). Hell, they could put an entire ARM computer system on the corner of one of their graphics cards if they felt like it.

Re:Why?

[Rockoon]

90C is the max, so clearly less is better than 90C.

60's are well below 90C.. quite a bit of headroom there.

Re:Why?

[Anonymous Coward]

He's even spoken at length about merging iOS concepts into the desktop OSX.

I've copied some text straight from the apple web site for the "Magic Trackpad" that make touch sound like you're no longer cool without it:

"The new Magic Trackpad is the first Multi-Touch trackpad designed to work with your Mac desktop computer."
"And it supports a full set of gestures, giving you a whole new way to control and interact with what’s on your screen."
"Magic Trackpad gives you a whole new way to control what’s on your Mac desktop computer. When you perform gestures, you actually interact with what’s on your screen. You feel closer to your content, and moving around feels completely natural."

That's still a far cry from touchscreen as the sole or primary input for a desktop computer, which is what he would have actually been referring to before. For what it's worth, I would personally like to see touchscreens in use for desktop and laptop systems, but only to complement a traditional keyboard and mouse/trackpad setup, not replace it.

Re:Why?

[Daniel Phillips]

My dual-Arm Xoom outpowers a fair number of the desktop machines I see in current use, both in processor power and graphics.

Re:Why?

[Anonymous Coward]

ITYM "discrete video card" not "discreet" - although the giant fans on many video cards these days make them very indiscreet :P

Re:Why?

[Anonymous Coward]

Dual-Core 1.6 to 2ghz, 4gb of ECC DDR3 1033mhz RAM... and only about 4 watts for a system (at the 1.6ghz speed).

That's going to be slower than an equivalent x86-based machine, though.

The thing is, it will be slower *but draw one tenth the power consumption*. I want ten of these on a board, with about eight times the processing power for the same power draw as an x86 solution.

...and an equivalent clocked x86 is hardly a barn-burner -- a 2ghz Core 2 (eg what used to be top of line line in mid-2006) or a comparable ULV i5 with Turbo is about the slowest machine I'd recommend to anyone these days for general use -- some of the things (video playback) can be pushed to a GPU, but hardly all of them, and if the 1ghz+ dual core ARM chips in today's tablets are any example, even just for heavy-ish web browsing or non-trivial document reading (large PDFs) there is a long way to go.

I mean, we're not talking coding workstations or gaming machines here - for a lot of stuff, even today's top end mobile processors (drawing 45W peak) are only marginal, that's why many folks still have desktops. But even for general use, ARM has not yet caught up with Intel.

Re:Why?

[Anonymous Coward]

Even games that aren't that graphics intensive, like Civilization, are very heavy on number crunching.

This bugs me. I was playing Civilization recently (Freeciv) on a low-powered (Atom) desktop, and the response time was becoming noticeably slow. I first played this game 20 years ago. Hardware has improved by orders of magnitude sine then, and the game mechanics haven't changed that much: they're still simple enough that I can comprehend them, and I think I could implement them (slowly) with pen and paper if I had to. The same applies for other recent versions of Civ (3, 4, ... haven't played 5 yet): except for the fancier graphics, the load on the CPU shouldn't be that much greater.

Okay, they're probably written in higher-level languages, or with other means of expensive abstraction - but does this really soak up the ~1000x improvement in hardware performance sine Civilization was released 20 years ago?

Re:Why?

[rsborg]

... it's maybe the ONE thing I agree with Steve Jobs about -- touch does NOT work as a viable input method for a desktop.

He may have said that at some point, but you should know by now that Apple changes the kool-aid they serve every so often. He's even spoken at length about merging iOS concepts into the desktop OSX.

[trackpad literature snipped]

You do realize that a trackpad is not "touch" as in touchscreen, right? I use a MacbookPro with a magic trackpad, it's a whole lot more efficient than a mouse at numerous things. With things like BetterTouchTool, I can do even more.

The problem with "touchscreen" on a desktop is gorilla arm. Even an iPad with a keyboard is not very fluid... there's a reason that the MacbookTouch doesn't exist. Mouse/trackpad with keyboard has worked so well for so long for a very good reason.

Re:Why?

[Anonymous Coward]

There's no reason a Cortex A7 dual core @ 800MHz wouldn't be able to handle both of the tasks you listed with ease.

This is an unwarranted assertion. When he's developing, he may be running test cases with his software which take an unknown (to you) amount of power. Similarly, his trading software may or may not be cpu-hungry. Unless you know what he's developing, and what trading software he's using, how can you say that a particular processor would be sufficient?

Re:Why?

[smash]

Seriously, what is the reason for having a desktop ARM computer? Power consumption? I don't think there's a very large market for people who will settle for tablet-like performance in order to save a few dollars a month at most on electricity compared to existing low power processors. People with power grid problems will want something that runs on a battery anyway, and a tablet/netbook makes more sense there.

The number of iPad and netbook owners out there disagree with you.

Fact is, for non-gamers (or, hell, even people who prefer to game on other devices), an iPad is more than fast enough. In fact, hell, it has the fastest growing game market at the moment as well.

People want: low cost, low heat, durability (no fan) and no viruses (this isn't inherently an ARM benefit, but you need to throw away your software library to get here, so its a contributing factor - if you're not running your old software, your don't need compatibility with it). Conversely, there are a lot of ARM programmers out there, and the ability to run ARM software is becoming more and more attractive.

Touch and ARM have nothing to do with each other - ARM was simply the best CPU for that form factor.

I don't think we'll see ARM on the desktop though - the desktop will simply cease to exist, except in niches where people genuinely need the CPU horsepower - and even for most of them i predict that massively multi-core ARM machines will eventually get enough power for most of them too - certainly enough that its not worth running a different software platform to the rest of us.

This isn't likely to happen in the next 6-12 months, but unless intel and AMD are very careful, it is going to happen sooner or later. If apple were to release an ARM powered laptop with 1.5 times or double the current Macbook Air's battery life, they'll sell like hotcakes. And then the rest of the PC market will try and play catch up.

Re:Why?

[smash]

If much of the software you want to use is say... already compiled for iPad and on the app store, then this isn't the case.

Don't think apple won't pull the pin on this, I suspect we'll see an ARM powered macbook air like device (perhaps re-using the name "Macbook" or "iBook" rather than Macbook air and Macbook Pro) in the near future.

Apple are big enough to create that market, and once they do the PC market will follow suit (Windows 8 running on ARM is confirmation of that in my eyes).

Re:Why?

[smash]

Also, don't forget, intel's 64 bit architecture was at least an order of magnitude more expensive than AMD's and had ZERO software already available. ARM certainly doesn't face that problem, its cheaper and has the app store, along with plenty of other ARM experienced coders working on non-apple platforms.

Re:Why?

[smash]

So because you can't program an arm tablet from itself, you think an ARM computer will be the same. Short sighted much?

Re:Why?

[smash]

How about a 64 core ARM cpu running at 1.ghz? Because i doubt we're THAT far from this sort of device, if ARM makes inroads into the traditional computer market.

Re:Why?

[smash]

I think by stating 10-15 years, you're being overly conservative. I give it 2-5 years. If someone (apple for example) releases an ARM powered laptop like the macbook air at half the price, with SSD and 3d video - people will go to the store, realise that why yes, it IS fast enough to do Facebook, email and internet browsing, and replace their old x86 machine (that is probably slower, as it runs off a mechanical disk) with it.

People are already trying to do this with tablets - the average joe doesn't WANT a computer - they just want some of the functionality that a computer provides.

Finding out what subset that is, and providing a device to do that in a reliable and maintenance free fashion is exactly what Apple attempted to do with the iPad. For some people it works. For others, not so much. It is, however a step in the right direction.

If you genuinely want a high powered desktop PC, sure, buy one. 99% of the public however, don't.

Re:Why?

[smash]

And yes, trackpad works (using one here). IOS concepts coming to OS X doesn't mean its going to be a complete clone. What works will come across. What isn't suitable, won't.

Re:Why?

[smash]

Thats because you're running unoptimized shitty software on it. You won't need to ask web sites to re-code for mobile/arm devices - if performance is insufficient and they lose hits because performance is too bad, that will be motive enough.

Re:Why?

[smash]

Exactly, mod parent up. I know several businesses here that are doing exactly this with dumb terminals. Replace dumb terminal with "employee's own iPad/netbook/etc" and i think this is where the "consumerisation of IT" is taking us. The only reason my own company isn't doing this is because we have half of our users in the middle of nowhere (mine sites) where getting a reliable 256kbit to the entire site is often a struggle.

Re:Why?

[Waffle Iron]

Thats because you're running unoptimized shitty software on it. .

Most of the software ever written is shitty and unoptimized. Where do you get the idea that that's going to change?

You won't need to ask web sites to re-code for mobile/arm devices - if performance is insufficient and they lose hits because performance is too bad, that will be motive enough.

The proliferation of bloated Flash content on the web is strong evidence that your theory is false. (For example: go to any restaurant chain to look at their menu. In theory, you want a couple of kilobytes of information to find the price of a few entrees. In reality, you get to wait half a minute while megabytes of animations download before you see even one hamburger basket.)

Even if your theory were valid, it would only apply to sites that have a large proportion of phone users. That has little correlation with business usage, which is largely in-house intranet stuff. That's generally even more shitty and unoptimized than average, and fixing it would usually cost a good deal more than the numbers I mentioned above.

Re:Why?

[sjames]

There's something to be said for a tiny machine with no fans. A lot of people barely get their PC above idle anyway. You don't need a 6 core 3 GHz CPU to check email and do some light web browsing.

Re:Why?

[Anonymous Coward]

What laptop is that? I want one!!

Re:Why?

[Anonymous Coward]

Some people have work to do. Jackasses like you keep fucking up everything that gets stuff done. (see gnome, kde, microsoft ribbon, vista, firefox)

Re:Why?

[ttong]

Of course you can program a tablet (of any architecture) from itself, Apple isn't the only vendor in the tablet market but they are the only vendor who forbids their customers to do so. It's just that when the OS fails to boot, it's impossible to flash new firmware without tethering it to a host device.

Re:Why?

[CAIMLAS]

Microsoft's intent has been for sometime to remove their dependence on other vendors. What makes you think anything more than a thin client will be needed in a couple years? That is quite obviously the direct MS is going with Windows 8. Windows 8 + an ARM thinclient + MS terminal service applications. All the better if they're Office 365 applications, right?

Re:Why?

[Anonymous Coward]

ONE thing I agree with Steve Jobs about -- touch does NOT work as a viable input method for a desktop.

But maybe a coffee-table top?

http://www.popularmechanics.com/technology/gadgets/news/4217348

Re:Why?

[SendBot]

You do realize that a trackpad is not "touch" as in touchscreen, right?

let's see... on one you are interacting with virtual objects in the device in a touchy way, using a device that emphasizes "touch" using the word "touch" in a literal way. On the other you are doing the exact same thing, but the point of contact is on the screen instead.

You realize that the people in this thread confusing touch interface with the specific implementation of a touch screen are not me, right?

Re:Why?

[david.given]

Which one, BTW? I'm looking for one and could use recommendations.

Re:Why?

[That Guy From Mrktng]

I think he was talking about the versatility that gives working in x86 and having a proper desktop ready to be tweaked the way you need it, is not about cpu achitectures, it's about i/o: having specialized peripherals, grunt power and options for the output. I'm in the same boat as him as I studied Advertising because I have always loved brands and the design itself of a logo and such. But when I was teenager and I got my PC I started to tweak it to the way I need it. Soon I was the PC guy of my neighborhood, even before I started the college and did lot of money then, guess how, fixing the same x86 borked windows ME, making data recoveries and selling parts. I probably have more money thanks to the hackability and the x86-CLONE idiosyncrasies than my marketspeak and design skills.

In a world where everyone have tablets you can't simply make the things he or I did, theres less motivation to learn and do since you're just a consumer of information not a generator because it's really a pain to force a tablet to do something thats not meant to be done tapping and pinching and flicking stuff. If you have 16 you can probably spend the entire afternoon flicking some "photoshops" but in pro/creative/production world one just don't simply ARM into mordor is not like you can't code in a tablet, I coded some really simple Java apps in my old Sonyericsson P900 had the free time - it was a pain - omfg they ran! - meh.

That said I can't wait for our ARM overlords in the desktop so I can finally offer simple, cheap and easy to learn appliances for people that is not going to mess around as we nerds do. When you learn about a new tech trend or related you don't just dismiss it because it does not suit YOUR needs, because probably fits perfectly for the 99.9999999999999999% of the world population, which for all intends and purposes it's more important than you or me or parent.

I'm all for the "one computer per human being" ARM is the way.

Re:Why?

[unixisc]

If you're talking Itanium, it's hardly the example I'd use, particularly since despite having some major OSs ported to it, it's being abandoned today left and right by Microsoft, Red Hat, Canonical & Oracle: even the Linux guys (except Debian) don't seem to want to do anything w/ it. I don't think ARM will click on the desktop, except if Apple makes it their Mac platform.

Re:Why?

[Reservoir Penguin]

Last week I had a taste of your low power computing. I had to use a substitute workstation with some dual core amd cpu from around 2008, it was something around 2.5 ghz and 3 gigs of RAM. Just using Outlook 2007, skype, Firefox and Excel nearly brought it ti it's knees. And you 800mhz dual core Atom is probably is even more underpowered.

Re:Why?

[badkarmadayaccount]

Make and model, if you would be so kind?

Re:Why?

[badkarmadayaccount]

Separation of interface and presentation surfaces means having a cursor. I think tablets are a fad because it's one layer of indirection too many, though standard disclaimers apply.

Re:Why?

[badkarmadayaccount]

CPU cache is a bottleneck.

Re:Why?

[Walter Carver]

I have an ARM based laptop.

Where can I get one like yours?

Archimedes

[rve]

It was on the desktop first. I was a kid, not terribly good with money, and it was expensive, so I just missed out on being an early adopter.

Yes, here is a link.

[Anonymous Coward]

http://www.macrumors.com/2011/09/23/support-for-marvell-quad-core-arm-chips-found-in-xcode/

http://arstechnica.com/apple/news/2011/09/support-for-quad-core-arm-cpu-shows-up-in-apples-xcode-but-why.ars

Macs and Hackintoshes.

Re:Yes, here is a link.

[Lunix Nutcase]

It's quite easy to figure out "why" they added that. They will probably being use the quad-core A9s in a future iPad.

Re:Yes, here is a link.

[smash]

Or a new "iBook" or "Macbook" (not pro/air).

When the desktop is superseded

[drinkypoo]

When you can get a quad core smartphone with a halfway decent GPU, who cares? The only real problem is the lack of memory. Dock your phone, use its display for status updates and compute on your TV... or monitor.

Re:When the desktop is superseded

[Lunix Nutcase]

The people who run tons of software that is x86 only and has no comparable ARM version? People who do work for which ARM is supremely under-powered even with a quad-core version? Even a low end i5 can blow away the fastest ARM processors. This quad-core version will close the gap some, but it will still be far noticeable less performant.

Re:When the desktop is superseded

[lkcl]

not if it's got 4gb of ECC DDR3 1066mhz RAM, it's not. the NuSmart 2816 is a 1.6 to 2ghz Dual-Core Cortex A9 with two versions - one 32-bit memory addressing and the other 64-bit. they're sampling, now. i'm working to get them plugged in to the EOMA initiative:
http://elinux.org/Embedded_Open_Modular_Architecture/PCMCIA
http://www.openhardwaresummit.org/forum/viewtopic.php?f=5&t=502

Re:When the desktop is superseded

[Bert64]

Sounds familiar...

Like the people who run tons of software that is (sparc|alpha|hppa|power|mips) only and has no comparable x86 version? People who do work for which x86 was supremely under-powered even with quad processors? Even a low end Alpha could blow away the fastest x86 processors.

History repeats itself, attack from below pushes the more powerful, more power hungry and more expensive architectures into small niches...

Re:When the desktop is superseded

[drinkypoo]

Sounds sexy, but I don't see it in a phone yet. is it? and with 4GB? If so, that would be fantastic.

Re:When the desktop is superseded

[Arlet]

Making the situation even worse is the fact that there is a complete lack of standardization on the ARM platform, especially for all the peripherals. But even for the core itself there are many different variants. This can be an advantage for embedded developers, because it gives you lots of choices.

For binary software vendors, it's a nightmare, because they would have to support all these different versions.

Re:When the desktop is superseded

[lkcl]

this is why China's ICT have added the 200 most-common x86 instructions as "emulated" but hardware-accelerated emulation to their upcoming Godson / Loongson 2G MIPS CPU. throw-away statistic about the 2G: even the single-core 1ghz 65nm version of the 2G has such a powerful Vector Processing Unit that it can do 1920x1080 MPEG decode at 100fps. and they plan to put 16 of these on one die! they're insane! :)

Re:When the desktop is superseded

[smash]

Other than games (which there are plenty of on iPad anyway) - the number of users who *actually* need the power of even a core 2 duo, when you take out the requirements of running windows, virus scanner, etc is perhaps 1% of the market.

They will continue to be served by high power desktops. The rest of us will have machines we carry around with us.

In netbooks

[X10]

I don't see why I should have an ARM processor in my desktop pc, but I would love to have one in my netbook. It would boost battery life from two hours to ten - same as my Galaxy Tab.

Re:In netbooks

[fast turtle]

Buy an AMD Fusion based laptop. Battery life exceeds 6 hours and can easily reach 8 hours of normal office use. With a bit of effort and thinking, they can actually exceed 10 hours of run time while sitll offering full compatibility with Win7 and all of your software

Re:In netbooks

[cyber-vandal]

Should've bought an Eee Pad Transformer then.

Re:In netbooks

[laffer1]

It depends which AMD Fusion CPU you buy. An E series might do this, but the faster A series does not. I just bought an A series a few weeks ago and it gets about 3 hours. It's an amazing CPU and faster than I expected for a quad core 1.4Ghz chip. Getting it to work with linux and BSD has been a real pain in the butt though.

Re:In netbooks

[badkarmadayaccount]

What kind of trouble did you run in to?

Raspberry Pi pricing

[Anonymous Coward]

Was talking about the Raspberry Pi to several Arduino guys last week at the NYC Makers faire. Most are convinced that they can only hit that price point because they are given free chips by the manufacturers. (This happens for education quite a bit i'm told.)

Thoughts?

Re:Raspberry Pi pricing

[sunderland56]

An Atom-based CPU + motherboard costs roughly $75. The Raspberry is supposed to be $25/$35 but is nowhere near as capable (no hard disk interface, fewer USB ports, slower graphics). The Atom is a fully usable, if a bit slow, desktop; the Raspberry is brilliant for what it is designed to do, but would not be usable as your one and only computer.

If Raspberry added the components to make it usable, and charged market price for the CPU, then it would be $75 - and would still be slower than the Atom. So, what exactly is the point of an ARM desktop?

Foaming at the mouth? Really?

[Anonymous Coward]

Methinks this post is laced with a bit of projection.

EOMA Initiative

[lkcl]

it's a long story, but i've been working to get ARM-powered desktop machines and laptops into the hands of free software developers for some time.

one of the key problems are that the chinese and taiwanese factories have absolutely no software expertise whatsoever. some guy decides he got caught out by the USA and UK Governments placing embargos and tariffs on imported clothes a couple years back: his business was affected, so he goes "i know, i'll diversify, i'll make tablets, those are popular". so off he goes, he gets supplied with a GPL-violating Android OS right from the word "go" by a limited number of Chinese ODMs who are having a really hard time keeping hold of their software engineers, and it just goes downhill from there.

the other problem is, as can be seen from the insane amount of money spent by the openpandora group, that case-work for laptops etc. can well be in excess of $100,000. that means that anything like the "pegatron netbook" has to be bought in volumes of 250,000 and above in order for the R&D costs to be amortised over a reasonable period.

this is where the EOMA initiative comes in: http://elinux.org/Embedded_Open_Modular_Architecture/PCMCIA

by reversing everything on its head, and getting free software developers a modular architecture which _could_ be dropped into a mass-volume product, the tables are turned: those Chinese Factories can be supplied *by us* - Free Software Developers - with a completed ready-to-ship OS.

so, yes there's a board which is available that is similar in size and function to the pandaboard, origen exynos board, beagleboard, IMX53QSB etc., but unlike those boards, by complying to the EOMA/PCMCIA Open Standard it would be possible to literally drop that hardware-software combination straight into a mass-volume product, with the development effort of the required motherboard being nothing more than a low-cost 2 to 4 layer board that even KiCAD, Eagle or gEDA could do.

one key part of this strategy is to leverage arduino-like boards, like the leafpad Maple:
http://elinux.org/Embedded_Open_Modular_Architecture/PCMCIA/MiniEngineeringBoard

anyway i think that's enough for one slashdot post. bit of background and some additional links, here:
http://www.openhardwaresummit.org/forum/viewtopic.php?f=5&t=502

Re:EOMA Initiative

[fast turtle]

I'd think that it would be more sensible to use an existing standard form factor such as M-Atx/Itx for the boards due to the tooling already being available. Simply put, the only issue you'd have is what internal connectors (sata/ide/floppy/firewire/usb) are needed along with the backplane ports, video, usb, ethernet being the most obvious. Then design the board to provide the needed connectivity and be done with it. The main advantage is not reinventing the damn wheel and getting a standard board into the hands of the devs as quickly as possible. Hell you could even go with surface mounting of the cpu and make it non removable/upgradable to avoid incompatibility issues. Includea a PCI-E slot for video or just use one of the AMD/Nvidia mobile GPU's as used in laptops to solve that issue and you'd have an ARM development machine in the hands of the devs within 6 months at the most.

Re:EOMA Initiative

[Doc Ruby]

Are there any existing ARM PC boards that can take the forthcoming Xilinx Zynq CPUs? Nothing fancy - I want to run an Android desktop on it while I port my PIC embedded industrial controller firmware to Zynq/Android/FPGA. Any PC I can swap a Zynq into now, or that will be available in 2012?

Re:EOMA Initiative

[Doc Ruby]

Are there any existing ARM PC boards that can take the forthcoming Xilinx Zynq CPUs? Nothing fancy - I want to run an Android desktop on it while I port my PIC embedded industrial controller firmware to Zynq/Android/FPGA. Any PC I can swap a Zynq into now, or that will be available in 2012?

Or rather, is anyone putting a Zynq into EOMA? What would it take for my HW lab to do so?

Re:EOMA Initiative

[Anonymous Coward]

GPL-violating Android OS

Looks like slander to me.

Re:EOMA Initiative

[lkcl]

ok - i'm pleased to see a response here: there are several points that are good, and some are, when you look closer, turn out to be unrealistic for mass-volume so-called "embedded" products.

1) the first is form-factor. this is great! yes, one of the options being considered is to have a standard Mini-ATX/ITX motherboard (into which an EOMA/PCMCIA CPU card can be plugged, at the back). there are several embedded companies that produce Mini ATX motherboards as standard, for their "modules", so it is not a new concept, it is in fact a proven one.

2) the second is connectors / interfaces. if you look right across the board at the very latest ARM processors coming out *right now*, you can count the number of Cortex A8 and Cortex A9 systems (as well as Marvell's "ARM-compatible" range of processors) that have PCI-e on the fingers of one hand.

i'll say that again.

the total number of modern ARM processors with even a 1x PCI-e interface is *below* 5 (five).

now there do exist some Cortex A8s (e.g. the OMAP35xx series) which have a HPI bus, onto which you could put a PCI-e "PHY" chip as it's called, but the total number of companies doing actual PCI-e "PHY" chips is, also, very very limited. typically, any company which has PCI-e PHY interface is a "Fabless Semi" company that gets bought up very very rapidly by the likes of Mentor Graphics, Synopsys and so on.

3) the third is the sheer overwhelming disparity between the ARM CPU's power consumption and the average PCI-e-based GPU's power consumption. the absolute ABSOLUTE lowest power consumption PCI-e-based GPU i could find is one from SiS, it's an older 65nm CMOS process, and if you ramp its speed down to the absolute lowest it will go without keeling over, it uses 6 watts. SIX watts!! you wanna connect a 6 watt GPU up to a 0.5 to 1.0 watt processor be my guest!

4) Multi-layer boards at ATX/ITX form-factor are expensive. if you have the CPU on-board the Motherboard (rather than being on a separate card), you then are forced to have the most complex part - the CPU-to-RAM interface - push up the number of layers required for the *whole* motherboard. by contrast, if you do the CPU-plus-RAM as a separate tiny, tiny board, just presenting its interfaces (SATA, ETH, USB, I2C, RGB/TTL etc.) via a simple connector (e.g. PCMCIA 68-pin) then you've just saved a fortune on the cost of the main motherboard because the main motherboard PCB can be done as an ultra-low-cost 4 or even if you're really lucky or a very good designer as a 2 layer board.

5) The power requirements of standard PCs are 10 to 200x larger than is actually needed! 500 to 1000 watts i mean for fuck's sake that's just insane. these ARM processors, the fastest most powerful one available on the market right now (sampling) is the NuSmart 2816, and that uses _two_ watts (shock horror) at 2ghz. wow big fucking deal. why on god's green earth would you want to match a 2 watt CPU with a 1000 watt Power Supply?? the entire motherboard would probably need a big resistor just to draw enough current in order to convince the PSU that nothing's wrong! i'm not joking about that - i'm dead serious.

6) The level of integration on these so-called ARM "embedded" CPUs is so high that it's really not worth the effort. present a USB bus, present an Ethernet port, present an SATA socket, along with an HDMI out and maybe even VGA, you're done! ship the damn product out the door, it cost you $35 to make! don't believe that price? just look at the cost of the RaspberryPi - it's doable. the irony is that for $35 of the "upgraded" RaspberryPi you can get an 800mhz Cortex A9 with an AML-8726-M (single core) for the same price. and the same size. credit-card-sized. you have to ask yourself: why would you _want_ to fit a credit-card-sized computer into a 12 x 15 x 8in "Desktop" case?? :) why not fit it into a 4in x 5in x 0.75in box, instead?

so, whilst on the face of it, fitting into the "standard" - i'm going to go further than that i'm going to

Re:EOMA Initiative

[Doc Ruby]

When it's in print it looks like libel, a more serious and costly offense, because it's being published to n people, rather than just spoken to one or a few.

Re:EOMA Initiative

[Anonymous Coward]

Get SBUS style expansion cards and build systems into a case around the size of the sparcstation ipx and I might be interested in something smaller than my nice 'legacy' Desktop case. However the reason I *HAVE* said case is because Iike being able to expand my computer. If I can't have a minimum of two and preferably 6 or more devices connected to an internal high speed bus, I am not a happy camper. I'm sure the number of people who agree with me are not in the millions, but I would imagine they dwarf your 'one hand'. The problem is there are certain minimum sizes to modern pc expansion boards, and you simply can't fit those into a smaller case. Especially when you start looking at video cards, the sizes shrunk for a while when 2D display technology matured, but as soon as 3d started to hit the table, board size began working it's way back out again.

I don't see that changing in the next few years. Perhaps in a decade 3d tech will be in the mature phase that 2d once hit and board sizes will drop to tiny thumbnail sized chips that you can run off a solar panel, but much like the american muscle car, they're still busy making them bigger, faster, stronger.

Re:EOMA Initiative

[Anonymous Coward]

What kind of desktop computer uses 500 to 1000 watts? Maybe somebody with a dual quad core setup, SLI'ed video cards, etc. But that user will never replace their setup with a 2 watt system. I have a more reasonable system with a Core 2 Duo and an ATI 4670 video card, and the last time I hooked up my Killawatt to measure the actual power consumption the whole power strip (monitor, computer, etc) drew 110 watts. It's always "on" but goes to sleep when idle for 30 minutes, so it's probably only awake for 6 hours a day on average which I calculated to be like $24 a year. In the grand scheme of things that just isn't enough of an electricity expense for me to care, much less to accept ARM-like performance.

Sure, I can run spreadsheets and edit documents on a tablet processor just fine, but it's never going to be as snappy as it would be on a full powered system. I think people underestimate the speed that the "average user" wants even if their just browsing facebook and writing an occasional paper.

Re:EOMA Initiative

[lkcl]

ok - it's too early to tell, yet, for the Zynq. you kinda get a feel for what these CPU manufacturers do. they go out and buy off-the-shelf "blobs" from fabless companies, slap together a few datasheets and see if anyone's interested: Xilinx is no exception here :) so, nobody expressed an interest in what they called the 7010, but plenty of people were interested in the 7030.

so anyway - i'm meeting with a xilinx rep next month, to get across that having this available in a beagle-board-like form-factor would be uhn... good!

second point is: do read what i wrote about "PCs" - there really isn't any point in sticking to the legacy PC form-factor, there really isn't. do you recall those IBM Mainframes where the engineer used to ask people to leave the room when they were being serviced? it was because this full-height cabinet had *one* PCB in it, right at the back, in the corner. to put an ARM-based motherboard into a standard desktop PC case would be something similar, but with about 100 to 1,000 times the cubic volume of the desktop PC case entirely wasted.

Re:EOMA Initiative

[Doc Ruby]

I'm an old friend of the Xilinx scientist who kicked off this Zynq project starting several years ago and has driven it to the market. They've done a lot more than buy a COTS ARM blob and fired off datasheets. I've been waiting for over two years for the final product to reach market after I heard it was finalized in engineering. I have expressed an interest, and indeed I believe my requirements helped influence the specs for the low end chip.

I'm not going for a "PC" as in form factor. I'm going for a PC as in device architecture: CPU, RAM, storage (HD/SSD), monitor (at least a 20x4 char LCD), keyboard and mouse option (USB), ethernet, other IO (RS-232/485), running an OS (Android). We design and build our own embedded PIC devices, including enclosures, and we'll switch to the one that works best in our industrial environment. But the point is that I want a PC with an ARM CPU that is a Zynq, running Android, that I can target for development the same way I target a desktop or server from my Eclipse IDE. If that means I have to build my own Zynq on a PCMCIA to use commodity HW for the rest, I will.

So how do I do that? Once I understand what I need to do, I can also look to Xilinx for some support in doing it. But until I understand the path, and then accept it's for me, I'm not going to bother my Xilinx connections with it.

Re:EOMA Initiative

[Anonymous Coward]

Wow great post. Just seems to highlight the possibilities though.

1) Desktop computing that doesn't use 1000 watt PSU. What's wrong with that trend? Talk about green computing. Solar powered computers? Awesome, let's do it.

2) Raspberry Pi allegedly does Quake III at 40fps. Holy crap. Who cares about discrete graphics, go the Intel route and add more cores. And there are 3D processors for smartphones (Power VR, etc). Just make a low power bus standard for pluggable graphics, and since you're advocating pluggable ram/CPU too, why not.

3) Or, on the other hand, got 500 watts? Plug in 200 processors, or pluggable CPU/RAM/SOCs. Who cares if ARMs don't have virtualization. Why be virtual when you can just plug in another System on a chip to the mother of all motherboards? Screw buying a huge rack, just have a big desktop case with the side open or some pluggable front interface, and buy little credit-card sized blades with one or more SSD Flash+CPU+RAM systems right in it. Oh yeah, give me a whole bunch of HDMI outs so I can see the raw output of 5-10 of them on a wall of cheap 150 dollar 1920x1080 monitors.

MiniBlades! That would be awesome.

Re:EOMA Initiative

[lkcl]

oh _good_! very very good to hear. keep in touch, ok? btw, the EOMA/PCMCIA spec, we're *not* going for PCMCIA electrical interoperability: it's re-use of the 68-pins, much like the Conditional Access Modules for satellite TV.

Re:EOMA Initiative

[lkcl]

http://www.codon.org.uk/~mjg59/android_tablets/

Re:EOMA Initiative

[lkcl]

http://lists.gpl-violations.org/pipermail/legal/2010-March/001872.html

Re:EOMA Initiative

[lkcl]

http://lists.gpl-violations.org/pipermail/legal/2010-March/001936.html

Re:EOMA Initiative

[lkcl]

you'll understand when peak oil hits, and you can't _get_ 110 watts of electricity.

There's a problem with ARM computing?

[GameboyRMH]

Nobody told me and I've been using my N900 the whole time with no problems. Why am I the last to learn these things!?

Re:There's a problem with ARM computing?

[RR]

I was aware of the N900. I still use an N800. I just didn't need an upgrade yet, and I was waiting for step 5 of the 5-step mass market process for Maemo (becoming concerned about bureaucratic interference when Meego became the company's strategy and they decided to dump Debian for Linux Foundation silliness), when the Elop Effect happened. So, no, I don't think the N900 was successful.

What's the problem with the TrimSlice?

[earls]

Developer only? What is that non-sense? The TrimSlice ships with Ubuntu ready to use. ~$200 for the feature set is a steal, IMO. Not happy without a Dell logo or something? What's the problem with the TrimSlice?

Re:What's the problem with the TrimSlice?

[Anonymous Coward]

I could go for a little slice of trim right about now~

Re:What's the problem with the TrimSlice?

[olesk]

Developer only? What is that non-sense? The TrimSlice ships with Ubuntu ready to use. ~$200 for the feature set is a steal, IMO. Not happy without a Dell logo or something? What's the problem with the TrimSlice?

I bought one last week from here http://trimslice.com/web/

Cheap, cool (as in not hot) and fairly reasonably priced for what you get. Waiting for mine right now...

Re:What's the problem with the TrimSlice?

[lkcl]

the first problem is that the cost of the whole system's components is actually, as the RaspberryPi shows, somewhere around the $40 mark. also, if you look closer at the Tegra 250 which is used in the TrimSlice, it doesn't have SATA-II. and as i mentioned in another post here, the number of modern ARM CPUs with PCI-e on the market is less than 5.

so you can't expand it (ARM systems aren't designed that way), and it happens not to have the _exact_ set of features which make it truly desirable to free software developers, which is why free software developers are plumping (reluctantly) for the OpenRD Ultimate, and wishing that there was something better.

right now, it's all a big big compromise. and that's why i started the EOMA initiative, so that those "compromises" do not impact on the development of a product: you can always go swap out the CPU card for something better when it comes along. http://elinux.org/Embedded_Open_Modular_Architecture/PCMCIA

there was a discussion of the features available (and desired) on debian-arm a few months back, let me give you a link to jump somewhere into the middle of that:
http://lists.debian.org/debian-arm/2011/08/msg00012.html

Re:What's the problem with the TrimSlice?

[Anonymous Coward]

The tegra2 which is the cpu on the trimslice is a nice attempt at multicore arms but has several deficiencies. It runs at relatively slow clock. 1Ghz just doesn't cut it for a desktop. The worst part though is the lack of NEON support(The ARM SSE2 equivalent). At least NVIDIA is learning from their mistakes. The next tegra series will have a higher clock(1.5-2 Ghz) and the NEON parts. That generation of arms will be the one which will overtake the x86 line. It's kind of like the days of transition from 2-stroke to 4-stroke engines

Re:What's the problem with the TrimSlice?

[couchslug]

"I could go for a little slice of trim right about now~"

Me too, everytime I read the Golden Girls Theme Troll.

Re:What's the problem with the TrimSlice?

[Anonymous Coward]

I think it's a little expensive at the moment; compare it, for instance, to the various atom based desktops at the moment. For instance, the 250G HD model is coming at 240 UKP + VAT. Compare this to the Aluetia T1 which is 240 UKP after VAT. Now, the trimslice may have better video capabilities, and it have lower Power usage (although not that much as the HD will take the same draw). But, set this against the more mature software environment for the intel based desktop.

I think that the trimslice is a valuable addition to the ecosystem and I may buy one anyway, but the price precludes a number of applications that I have in mind. It will be interesting to see what effect the Raspbury PI has; if it results in low cost computing in the way that I think OLPC pushed the netbook market, then I think that will be a notable victory.

Search a little more, like the Efika

[jonsmirl]

http://www.genesi-usa.com/products/efika
Smarttop $129 thin client
Smartbook $199 laptop

They run Ubuntu and are based on the Freescale iMX51.
They are far more powerful than a Raspberry PI.

Freescale i.MX515 (ARM Cortex-A8 800MHz)
3D Graphics Processing Unit
WXGA display support (HDMI)
Multi-format HD video decoder and D1 video encoder (currently not supported by the included software)
512MB RAM
8GB Internal SSD
10/100Mbit/s Ethernet
802.11 b/g/n WiFi
SDHC card reader
2 x USB 2.0 ports
Audio jack for headset
Built-in speaker

10.1" TFT-LCD, 16:9 with LED backlight, 1024 x 600 resolution
Freescale i.MX515 (ARM Cortex-A8 800MHz)
3D Graphics Processing Unit
Multi-format High-Definition hardware video decoder
16GB Nand Flash
External MMC / SD card slot (up to SD v2.0 and MMC v4.2)
Internal MicroSD slot
802.11 b/g/n WiFi (with on/off switch)
Bluetooth 2.1 + EDR
2 x USB 2.0 ports
Phone jack for headset
Built-in 1.3MP video camera
Built-in microphone
Built-in stereo speaker

Re:Search a little more, like the Efika

[Anonymous Coward]

Panda board is likely decent enough for a desktop. (Use hard fp debian port).

Re:Search a little more, like the Efika

[Anonymous Coward]

Captain Anonymous would like to swing by and thank you for bringing this to his attention. He has never seen this before and is now thinking of buying one. It should keep him happy until he gets a Raspberry Pi.

Re:Search a little more, like the Efika

[fnj]

Don't get me wrong; it's good that somebody's trying, and we need a LOT more people trying, but jeeze. Looks like a two big giant misses to me.

The desktop has HDMI yet it can only display 1280x768? WTF man!!! There is no conceivable excuse whatsoever for no 1080p. No gigabit ethernet? WTF! Only 2 USB ports? WTF! OK, if it was like $50 all that might be acceptable, but not for $129. Not even close.

The notebook is only 1024x600. That's completely unacceptable to me, but hey, there are plenty of other Netbooks that get by with that much, so I'm in the minority. So ... battery life is 6 hr (claimed). WTF! You can get any kind of crappy x86 netbook with a claimed battery life of 6 hr. And for not much more money either. So what the hell is the point of this thing? If the battery life was 24 hr claimed, or let's say AT LEAST 16 hr in REAL LIFE actual use, it would have my attention, but as it is, it's just one more piece of junk.

Re:Search a little more, like the Efika

[e40]

We're on our 3rd MX Smarttop in a month. The first 2 kept turning themselves off after a few hours to a day. We have a ton of equipment in that server room, and nothing else has power issues, so it is not our power. It's their quality. The 3rd one seems to be OK, but beware. Several of us were thinking of buying them for home, but I'm not after the quality issues.

Obligatory...

[cjjjer]

2012 will be the year of ARM on the desktop...

ARM desktop from the past

[Anonymous Coward]

I already have one (Archimedes) lurking around in the basement (http://en.wikipedia.org/wiki/Acorn_Archimedes)

Here's a couple:

[Roman Mamedov]

http://www.genesi-usa.com/products/efika
http://trimslice.com/

Re:Here's a couple:

[lkcl]

the trimslice is $200, is non-upgradeable and does not have an SATA-II interface (whereas the raspberrypi is $25 - a discrepancy of 85%+ in price, for no good reason). the genesi products use an iMX515 which has a hard limit of 512mb RAM, and the laptop only has a maximum LCD resolution of 1024x600, which is completely unusable. every single product out there right now has these hard compromises that make them completely dissatisfactory for at least one market. this is why i'm doing the EOMA/PCMCIA initiative, so that a product is *not* restricted (requires a total motherboard redesign) to a particular CPU. http://elinux.org/Embedded_Open_Modular_Architecture/PCMCIA

There is a smartbook available (although not good)

[IYagami]

The Toshiba AC100

You can find a review at http://www.reghardware.com/2010/11/03/review_netbook_toshiba_ac100/

"The beautifully designed and executed hardware is very close to my ideal netbook, and it's hardly an exaggeration to say that I'm heart-broken by Toshiba's cocked-up Android implementation. The best one can hope for is a firmware rescue from the open source community, although I wonder if the product will stay around long enough in these tablet-obsessed times for that to happen."

Has Everyone Already Forgotten About the Netwinder

[LuxuryYacht]

The NetWinder was based on the DEC/Intel StrongARM 110. They had quite a nice desktop working back in 1999 along with a large developer community

Yes, in about two months.

[Anonymous Coward]

Raspberry Pi. $35 for a 700MHz ARM with 256MB of RAM and 1080p HDMI output. More computer than most people need.

Re:Yes, in about two months.

[Anonymous Coward]

Isn't the Raspberry Pi the device with the Broadcom ARM soc without any public docs? The Raspberry Pi director does work for Broadcom. Does Broadcom even have a single ARM device with open docs?

Re:Yes, in about two months.

[LuxuryYacht]

The Broadcom BCM2835 looks like the ARM SOC used by the Raspberry Pi. There isn't a link for its data sheet that I could find. So it looks like just another closed hardware ARM design.

Re:Yes, in about two months.

[Anonymous Coward]

Who cares, as long as the kernel drivers and any necessary userspace tools are open source? Does it make a difference if the graphics card loads its firmware from a flash chip on the card or a binary blob on the SD-card? What documentation do you need?

Re:Yes, in about two months.

[Anonymous Coward]

Malware in the closed binary blob firmware and bootloader. Closed gpu libs. Not being able to boot your own kernel. Not being able to boot off the block device or location of choice. To name a few.

Lots of people care. That is why so many people reverse engineer their ARM tablets and cell phones. That is why there are open firmware projects like http://www.coreboot.org/ coreboot, http://kexecboot.org/ kexecboot and http://www.denx.de/wiki/U-Boot u-boot. Take a look at the http://wiki.opengraphics.org/tiki-index.php Open Graphics Project or http://elinux.org/Embedded_Open_Modular_Architecture/PCMCIA mentioned earlier in the comments here for what and why.

Re:Yes, in about two months.

[Anonymous Coward]

Malware can be in the flash memory on a graphics card. I don't see people complaining about closed source firmware there.

I said open source kernel drivers and user space utilities, so "closed gpu libs" and "not being able to boot your own kernel" are out.

Regarding "not being able to boot off the block device or location of choice": It works like most other embedded devices. There's some flash memory which gets read first, containing a kernel or a boot loader. It just so happens that in this case the flash memory is an SD card. Unlike other devices, you can't brick this one by flashing a bad/wrong image, because you can always just put the SD card in another computer and fix it. You can load your actual operating system from any device you like, provided you compile a suitable boot loader. The embedded world has quite a few to choose from. You mentioned some of them.

If everything works to plan, everything that runs on the ARM core will be open source. The only piece of proprietary software on that system will be the GPU firmware. That blob will load first, into the GPU, and then your kernel or boot loader. From your software's point of view, it's as if the system came with built-in firmware - firmware that would never be touched or whined about if it were the firmware of a PCIe graphics card.

Re:Yes, in about two months.

[Anonymous Coward]

Sure $35 without a case or power supply, and you still need a display, keyboard and mouse.

The $100 XO 1.75 looks like a much better deal. http://www.engadget.com/2011/01/09/marvell-powered-olpc-xo-1-75-only-draws-2-watts-of-power-finall/

It even comes with a touchscreen display, wifi, USB, keyboard, battery and handcrank. http://wiki.laptop.org/go/XO-1.75

Before you go saying that ARM is fast enough...

[CajunArson]

OK, the idea behind ARM is that it is "fast enough" for desktop and notebook PCs. Well, if that's the case, then a P4 is also "fast enough" and you should consider not buying anything newer.

Why am I saying that? Let's look at one benchmark that *is* multi-core ready and that Nvidia kindly ran on the upcoming Kal-El quad-core systems: Linpack.

Now I know Linpack is not a perfect benchmark, but it does do a decent job of showing off number-crunching power and it is multi-core capable and there are results from a wide range of architectures.

Here's a result from a 1.7 Ghz P4 system (see: http://www.roylongbottom.org.uk/linpack%20results.htm)

CPU Mhz Opt (MFlops) Non-Opt (MFlops)
Pentium 4 1700 382.00 131.59

I think (but I'm not sure) that Opt means optimized (such as using SSE) and non-Opt is a minimal x86 implementation with no optimizations.

Now, here are Nvidia's results for its not-yet-on-the-market Kal-El Quad Core ARM at 1.0 Ghz:

Multi-threaded Linpack: 309 Mflops

See: http://www.xbitlabs.com/news/mobile/display/20110921142759_Nvidia_Unwraps_Performance_Benchmarks_of_Tegra_3_Kal_El.html

I'm going to assume that Nvidia will go out of its way to make sure the code is optimized for benchmarks that it posts as part of a marketing push.

So a QUAD CORE Arm architecture is still lagging behind a P4, and while the P4 has a clock speed advantage, it's a lot smaller than is justified by the difference in performance considering the Nvidia chip has 4 cores compared to a single-core P4.

Now, I'm not saying that Kal-El won't be awesome for use on tablets and smaller devices, but on a desktop or even a notebook, don't go around expecting miraculous performance.

Re:Before you go saying that ARM is fast enough...

[CajunArson]

P.S. --> Since Slashdot formatting is whack, the above results should show the 1.7 Ghz P4 with a 382 Mflop Linpack score when "optimized" compared to Nvidia's published results of 309 Mflops.

Mathematically, the P4 has a 70% clockspeed advantage with only a 23.6% performance advantage, but remember how crappy the IPC on the old P4 was, and remember that the P4 is a single core CPU vs. Quad Cores for Nvidia.

Re:Before you go saying that ARM is fast enough...

[statusbar]

How much power does the 1.7 Ghz P4 system require vs the Kal-El Quad Core ARM at 1.0 Ghz?

Re:Before you go saying that ARM is fast enough...

[Anonymous Coward]

It's useless to ask; he can't hear you because of the cpu-fan.

Re:Before you go saying that ARM is fast enough...

[DuckDodgers]

2 points to consider:
1. The P4 uses something like 5-10 times as much power for the task. Considering that these days most people just surf the web and play simple games that can be programmed with Javascript and HTML5, that means ARM is good enough for a large minority of current PC users. It also means ARM is pretty good for a lot of laptops. At our company the software developers run IDEs and compiler on laptops, so beefy x86 CPUs are extremely useful. But everyone else is using a web browser and Microsoft Office - as long as the user interface remained the same, a switch to ARM wouldn't even be noticed.
2. ARM processors are going to improve. The nVidia Tegra 3 "Kal El" processor is made with a 40nm semiconductor technology, 4 cores plus one low power core, a cap of 1.5 GHz, and a 12 core GPU. nVidia has announced future generations of its ARM processors with 28 nm semiconductor technology, 8 cores and a 64 core CPU, and no details on max GHz. It's likely in as little as three years there will be ARM chips that are comparable to a mid range Intel Core i5 today with much lower power draw. Then ARM will be sufficient for 75-90% of desktop machines.

I don't do video editing, Folding@Home, or Bitcoin mining. I'm sure if I did that would change my position on this, and for people that do I'm sure having a beefy CPU and top tier dedicated GPU is essential. In my experience, the AMD 6-core CPU I have at home and the Core i5 I have at work never top 100% use of one core plus 50% use of a second core. The rest are always idle.

Re:Before you go saying that ARM is fast enough...

[Anonymous Coward]

How much power does the 1.7 Ghz P4 system require vs the Kal-El Quad Core ARM at 1.0 Ghz?

Does it matter?? Seriously, he was comparing old tech with new tech. If you want power comparison, compare at least somewhat modern tech with teh Kal-El.

Let's take the SINGLE core AMD Sempron 145 @ 2.7GHz with 45W TPD. It will clearly blow the P4 and Kal-El out of the water in performance. But you can even compare the power usage. At idle, it clocks down to 800MHz and uses under 2W power.

For even lower power APUs (ie. CPU + Radeon cores)

        http://en.wikipedia.org/wiki/Bobcat_(processor)

Intel is compatible with power usage at idle. There is no reason to go to ARM to lower power output of desktops and laptops.

Re:Before you go saying that ARM is fast enough...

[Anonymous Coward]

Kal-el @ 1Ghz 1.261W
Pentium 4 @ 1.7Ghz 64W

Kal-el is also including the power for the graphics, memory controller and a whole bunch of other stuff normally done by P4 chipset.

Re:Before you go saying that ARM is fast enough...

[Anonymous Coward]

I've always found pure number crunching benchmarks useless for anything but knowing how fast my povray images might render :-\

Memory bandwidth is *extremely* important for real world tasks. Do you have a graph for a memory benchmark? I have a feeling that the arm will be completely blown away.

Re:Before you go saying that ARM is fast enough...

[jon3k]

The Cortex-A15 will be available in up to octo-core configurations at 2.5Ghz, using a fraction of the power of a P4 (vs the 1.0Ghz benchmark you provide). You can make up arbitrary hardware requirements for "most people" but that doesn't exactly explain why Apple is selling tens of millions of iPads to ecstatic customers with a fraction of the "power" of a P4.

Re:Before you go saying that ARM is fast enough...

[Anonymous Coward]

According to the benchmark you posted, a dual core Kal-El hit 200 MFLOPS. Since Linpack scales almost perfectly, the quad core Kal-El should have done 400 MFLOPS, but only does 309 MFLOPS. 400 MFLOPS is faster than that old P4; it looks like the main reason for the lagging performance is the memory system, not the CPU cores. Another post mentioned an eight core system, which (with perfect scaling) you'd expect to get 800 MFLOPS on, which easily blows that old P4 out of the water.

Re:Before you go saying that ARM is fast enough...

[Anonymous Coward]

Is anyone predicting performance miracles?

The power draw differential between the SOC and a Netburst chip. Where are the numbers on that? It would be hilarious.

That chip on a decent flash storage SSD would appear faster and more responsive than a P4 with disk storage from that era. We forget how much everyday computing is IO bound. Gaming is all hot-cache computing.

Re:Before you go saying that ARM is fast enough...

[Billly Gates]

INT and FPU performance in MFlops does not make a computer faster.

For a consumer, it means to turn on their pcs fast, have the web smoothly scroll up and down like a Droid or IPhone, not flicker like a PC. They want their icons to open and close smoothly and want fast quick MS reaction times when completing commands.

A better GPU and quick reaction time from flash ram can do this better than a great CPU that has aesome MFLops. My 700 mhz phone can browse slashdot.org better than a 3.0 ghz hyperthreaded Dell at work! Why is that?

It is because the ARM has integrated graphics which are better than the crappy drivers/hardware integrated in that much faster POS Dell. Add to the fact that the browsers except for IE 9/10 do not accelerate video and it shows. Even in Chrome you can turn on accelerate Canvas and slashdot always flickers when you hit the up and down arrow keys compared to an IPAD or Driod. IE 9 on my computer is somewhat smooth and better but it has an ATI 5750. Again it is the GPU but still it is not as fast nor smooth as an IPAD which shows bad drivers.

Chrome and Firefox need to accelerate more HTML rendering to the GPU still as it is only partially accelerated.

The Llamo and AMD chipsets will appear much faster than anything by Intel just because they want quick animations to them makes quick and better performance. But in reality x86 is playing catchup to the arm in this area.

The rise of the tablets are showing how bad Windows and x86 really are. Things need to improve and catch up as to the consumer the x86 is quite slower and bulkier and more expensive than a tablet.

Mflops might be quicker if you compile code or do simulations or the things slashdotters do, but for general experience the age of x86 and the PC architecture show.

Re:Before you go saying that ARM is fast enough...

[evilviper]

The Cortex-A15 will be available in up to octo-core configurations at 2.5Ghz, using a fraction of the power of a P4 (vs the 1.0Ghz benchmark you provide).

MHz myth in full force... Don't we still kick people off of /. for stupid mistakes like this? In short, that 2.5GHz could well be slower than the 1GHz CPU. Things like pipeline length, bus and memory speeds, and MIPS per MHz are important.

Additionally, he mentioned the SLOWEST P4 he could find... They're basically giving away much faster P4s these days. Hell, you can buy a complete 2.4GHz P4 system for under $40: http://www.geeks.com/details.asp?invtid=SAMBA845V-24-4-R&cat=SYS

Even cheap and useless gimmicky junk like the Qi don't compare (where's the video? where's the optical drive? Where's the expansion?)

You can make up arbitrary hardware requirements for "most people" but that doesn't exactly explain why Apple is selling tens of millions of iPads to ecstatic customers with a fraction of the "power" of a P4.

Come on, this is PAINFULLY obvious...

The iPad doesn't run Mac OS, RHEL, or Windows. It runs code that was optimized to hell and back to be small and fast. I've got my old Psion 5MX with a 26MHz ARM CPU that is more responsive than an iPad, but that doesn't mean that ancient 26MHz ARM CPU is on-par, performance wise.

Also notable is the affect of solid state. I tried an Archos 70 with a 256GB laptop hard drive in it... It was immediately PAINFUL to use. The fact that tablets / smart phones come with solid-state storage helps to mask a lot of the poor performance, so the lack of IO lag makes it FEEL like a lot faster.

People have accepted crippled (fast) software on tablets. Turn it into a desktop, however, and they'll be feeling the limitations very quickly, and get very upset. And if you take away the limitations (eg. install Linux on it) you'll find yourself feeling the painful performance limitations very quickly.

Re:Before you go saying that ARM is fast enough...

[evilviper]

How much power does the 1.7 Ghz P4 system require vs the Kal-El Quad Core ARM at 1.0 Ghz?

Actually, and old P4 will idle down to pretty damn low power. But that really doesn't matter... the CPU will be only a tiny fraction of the overall power consumption.

I recently bought an Athon x4 with a TDP of 45W. At idle, it's using next to no power... Yet this super-efficient sytem draws 65 watts at idle. Meanwhile, the ancient P4 I'm typing on right now, is idling at under 40 watts. WHY, you ask? Because as memory bandwidth increases, Northbridges draw vastly more power. The memory itself feels nice and cool, but the northbridge needs its own fan... Putting an ARM CPU in a desktop motherboard would NOT allow it to avoid these problems. Sure, you could go with nice slow buses (ala phones, tablets, pdas, etc.) but then you'll have a horribly performing device.

And did I mention video cards? No big deal when you're single-tasking on a tablet, maybe with a game using a tiny subset of OpenGL commands, but go to a desktop, and you won't even be able to support the native resolutions of an LCD monitor (I haven't seen a phone that outputs even 1080 on their HDMI, yet). So now your nice sub-watt CPU is drawing 60 watts for fast memory bus, and a graphics chip idling.

Re:Before you go saying that ARM is fast enough...

[evilviper]

1. The P4 uses something like 5-10 times as much power for the task.

The power consumption of an idle P4 is just noise. The rest of the system (fast northbridge, GPU, etc.) consume VASTLY more power, so 5-10 times as much power is likely to be well under a 1% difference in total.

But everyone else is using a web browser and Microsoft Office - as long as the user interface remained the same, a switch to ARM wouldn't even be noticed.

WRONG! Your phone and tablets don't feel slow because the software was custom written to be extremely fast. Your web browser is actually extremely CPU-hungry, you just don't notice because you've got a system that's vastly faster than a phone... Have you ever tried running Firefox 3.x on a 1GHz Duron, with PC-133 memory? That was my primary workstation until recently, and it was MASSIVELY slow. Firefox took FOREVER to open a page, while something feature-bare like links-gui or Dillo was damn near instant.

And BTW, Firefox is a lightweight in comparison to Flash. One visit to Youtube and after being unable to play more than a couple frames of a video and your ultra-efficient ARM Desktop system will feel like the 33MHz 486 it really is...

Re:Before you go saying that ARM is fast enough...

[jon3k]

MHz myth in full force.

Except that ARM already confirmed it's 40% faster, core for core, clock for clock. I'm pointing out not only is it a newer and higher performing micro-arch, it's _ALSO_ a clock speed increase. The A15 will blow the A9 out of the water.

The iPad doesn't run Mac OS, RHEL, or Windows.

And why do you assume that an ARM based desktop should run the same software as traditional x86 desktops?

Re:Before you go saying that ARM is fast enough...

[Anonymous Coward]

"Before you go saying that ARM is fast enough... (Score:3) by CajunArson (465943)"

Someone give this fella a +1!!!

Re:Before you go saying that ARM is fast enough...

[evilviper]

I'm pointing out not only is it a newer and higher performing micro-arch, it's _ALSO_ a clock speed increase. The A15 will blow the A9 out of the water.

Fair enough, but those are marketing numbers, I'm not inclined to believe them for a second. When 3rd parties have their hands on them, we can talk. And as I keep saying, CPU is only part of the equation. If they don't have very fast buses and memory, and better GPUs like desktop systems (driving up power consumption for sure) they will still perform quite poorly.

Re:Before you go saying that ARM is fast enough...

[jon3k]

While I absolutely agree that an ARM based system using current (and even the next gen) of ARM reference designs would NOT be suitably for 100% of x86 users today, I think it's entirely possible that the next generation (A15 and beyond), with the appropriate software, could satisfy the computing requirements of a lot of people. And what about a generation after that? I think you overestimate the performance requirements of the general population. Most people don't post on slashdot, and don't have our requirements.

(driving up power consumption for sure)

Not necessarily, not if they also reduce the die size of the fabrication process. You could actually have higher performance at the same power consumption.

Re:Before you go saying that ARM is fast enough...

[evilviper]

I think you overestimate the performance requirements of the general population.

This goes back to what the GP said... If you're okay with poor performance, why not buy an old P4 system for $40 (or keep the one you have)? If ARM is perpetually behind x86, what is the advantage to ARM?

Low CPU power consumption is a joke in a desktop system, because all the other power-hungry components will mask any savings. It's possible to get impressively low-power x86 CPUs, like a Athlon x4 2.5GHz that maxes out at 45watts (idles in the single-digits).

ARM obviously won't run pre-existing desktop software, and anybody that's run a Linux desktop on non-x86 boxes will tell you that it's a decided downgrade, as numerous porting issues crop-up. The lack of Flash would of course be an issue, and no, normal/average people won't go without their YouTube, gnash isn't even remotely close, and HTML5 is a long ways off.

I'm still waiting to hear the advantage. Price is higher, power consumption difference is negligable, performance is considerably lower, etc.

Re:Before you go saying that ARM is fast enough...

[jon3k]

1. An old P4 uses many many times the power.
2. Lower power for desktop is good when your "desktop" is the smartphone in your pocket docked to a workstation.
3. ARM wouldn't need to run pre-existing desktop software, just mobile software like android ("apps" + browser)
4. Youtube works fine on an iPad or Android device today and you're right I'd agree people can't live without it.
5. Buying a used P4 for $40 isn't really a fair comparison to buying a brand new ARM chip using a fraction of the power. Compare new vs new or used vs used, and keep the wattage the same. Really it comes down to MIPS/watt and the upper bound needs to be incredibly low.

Re:Before you go saying that ARM is fast enough...

[DuckDodgers]

For power consumption, I'm speaking of power use under load. For laptops especially this is important. I'm sure that under idle the difference between a P4 and an ARM desktop is smaller - I'm sure it adds up over time, but not enough to be noteworthy.

With respect to performance, I'm not suggesting that my coworkers replace their Core 2 Duo laptop processors and Core i5 laptop processors with the single core 1 GHz ARM chip in my wife's smart phone or even the dual core ARM chip in an iPad 2. I'm suggesting they be replaced with an nVidia Tegra 3 "Kal El" chip with four core ARM processor at up to 1.5 GHz or the following generation eight core ARM processor with up to 2.5 GHz that nVidia announced. It still might not perform as well as the Core 2 Duo, but most of the time they're waiting for things on their device it's due to the SATA hard drives or website response times and not the processor.

Re:Before you go saying that ARM is fast enough...

[evilviper]

1. No it doesn't, as I keep saying.
2. In this case, you don't want a desktop and are completely off the topic.
3. Mobile software works for mobile devices. Once you give people a computer interface, they're going to want to do real work on it, and being unable to do so won't be acceptable as it is now when people have a mobile phone + tablet + computer.
4. Youtube works on phones because of proprietary applications that interface directly with it, blessed by Google. If google decides they don't like your desktop idea, you won't get your youtube.
5. YES IT IS FAIR because you're suggesting a NEW ARM CPU will be performance-competitive with said old/used P4 CPU. You certainly can't claim anything ARM will be REMOTELY competitive with a NEW Intel/AMD CPU. This is GP's original point, and I can't understand how you can be so incredibly dense as to not be able to understand the concept, time after time...

Re:There is a smartbook available (although not go

[cyber-vandal]

Or the Eee Pad Transformer which by all accounts is very good

The OP is talking about red apples, not greenapplz

[SendBot]

The OP said "touch interface" and then refers to "touch" as an "input method". The trackpad is described as "multi-touch".

I stand by my statements and provide further evidence of "touch" making its way to the desktop. Look! It's an official apple support page about multi-touch gestures in OSX Lion, one of the big things they were promoting about it: http://support.apple.com/kb/HT4721

If you're touching a trackpad (distinct from mousing, which is more using a stick to poke things instead of touching directly), what and where you are touching the interface is largely arbitrary. It's not that crazy to imagine a kinect-like desktop interface being common so that people can touch items in their desktop experience without smudging the screen up with fingerprints.

windows 8

[CobaltHex]

well windows 8 will be on arm

Re:windows 8

[gl4ss]

meh.

metro runs on arm and x86. ..in a different os, that they'll also just happen to call windows 8.

windows has been "running on arm" for a decade. http://www.dealextreme.com/p/7-tft-lcd-windows-ce-6-0-arm-cpu-wifi-umpc-netbook-266mhz-2gb-flash-disk-usb-host-sd-slot-lan-39391 80 bucks. windowed windows, on arm.

Wrong subject

[Anonymous Coward]

I think people are looking at this completely wrong. How much does a ARM device cost? 600? 800$? 1200? most of that is the battery, screen and maybe the flash itself. When you look at the AppleTV (99$) you notice none of those parts are a consideration.

So how do you get everyone on the internet with their own identity? Easy, make the computer core portable (the CPU, OS, and primary storage) that you can just plug into any other computing device (eg iPhone, desktop.) When the next more powerful version comes out, you migrate it to the new one, without having to throw away all the more expensive parts (Screen, batteries and secondary storage.)

Of course this would cut into the profits of all the companies involved, but from a cost and ecological point of view it makes sense. Instead of a 1000$ desktop or laptop, you have a 99$ micro-desktop that you can take anywhere and plug into anything. "The cloud" promises this somewhat, but is unworkable if you're in places without the internet, or in countries that eavesdrop on everything you do. It largely solves the malware problem since everything you put on the portable module is yours, and you don't have to constantly authorize and deauthorize each computer you work at.

Anyway. ARM desktop, ideally would be a 2" by 1" module that you can insert into a iMac-like computer screen to get a desktop, or iPhone-like portable to get a laptop/smartphone. No syncing, no batteries, no wireless, no passwords(unless you set one up,) all the hassle cut out, because the CPU,RAM,primary storage and OS are on the module. We're probably 2 years away from this being actually viable, mainly the cost of flash and the lack of appropriate GPU performance is what's holding things back.

Seeing an ARM desktop that's like a full x86 desktop in a ATX formfactor is a non-starter as that eliminates two of ARM's advantages - power and size. Intel has to cripple their CPU's down to the performance of a 10 year old computer to get the power envelope to ARM's capability, and ARM still winds up being about as powerful as a 5 year old computer. So ARM has at least a 2:1 performance to power benefit.

Re:Wrong subject

[gl4ss]

if you go with shitty via socs you can get functional "computers" for under a hundred bucks.
dealextreme has several models, with keyboards or only with touch. what makes the latest generation arms cost so much is that there's more demand than output, because they're latest gen, and orders made in advance. for general computing they already fill everything - but not for something that you need somewhat new computer for, like creating graphics etc.

http://www.dealextreme.com/p/7-lcd-windows-ce-6-0-via8650-cpu-umpc-netbook-w-wifi-pink-arm-v5-349-79mhz-2gb-sd-lan-82674

Acorn Archimedes - the first desktop ARM

[Nivag064]

ARM got to the desktop years ago (1987, according to wikipedia), as the first computer to use the ARM chip was a desktop computer - the Acorn Archimedes!

I had one, it was a lovely computer easy to program, and a GUI for in advance of its time.

Zynq-7000 PC?

[Doc Ruby]

What I want is a desktop Zynq-7000, the ARM A-9 CPU from Xilinx with a large embedded FPGA, running Android. My lab desktop, anyway. I want to port my embedded industrial control PIC code to it, perhaps targeting a soft PIC core in the FPGA (at first, then gradually porting sequential functions to Android processes). A desktop ARM/FPGA would be a great way to use the large universe of desktop apps to get the embedded PC to do what I want, even if I then repackage it as an embedded device (text LCD, minimum IO ports, no local Desktop, etc).

The Zynq CPU itself is probably not shipping until 2012. But who's got a PC built on it in the pipeline? Who's got some other ARM PC that could take a Zynq popped into it with a minimum of electronics work?

Re:Zynq-7000 PC?

[lkcl]

yes - and it gets _particularly_ interesting if there is a free software toolchain for Xilinx FPGAs. oh wait - there is one! http://www.milkymist.org/fpgatools/

Re:Zynq-7000 PC?

[lkcl]

imagine having a Zynq-7030 as a desktop PC in a PCMCIA-sized unit, and on the end of the 68-pin connector there was an Arduino-compatible 2in x 2.5in Leafpad "Maple"-like Embedded Controller as well (STM32F). would that absolutely rock, or would that absolutely rock? oh wait - this is what's possible as outlined by the EOMA/PCMCIA initiative, here: http://elinux.org/Embedded_Open_Modular_Architecture/PCMCIA/MiniEngineeringBoard

Re:Zynq-7000 PC?

[Doc Ruby]

What I want (for myself and for a productive community) is tools for Eclipse. So I can create a single project containing both CPU and FPGA code/resources that the IDE manages, spawning tools per target respectively. An integrated single-stepping debugger (integrating FPGA states into the hybrid CPU environment) would be ideal. But any IDE integration would be a step. Are there ways to package those tools for Eclipse to use them? Their wikis to which you linked don't indicate so.

Re:Zynq-7000 PC?

[Doc Ruby]

Why just imagine, when I have a lab that maybe could do it? As I asked in another post in this discussion, is anyone putting a Zynq into EOMA? What would it take for my HW lab to do so?

Re:The OP is talking about red apples, not greenap

[Sancho]

Smudging the screen isn't the problem. The problem is holding your arm up for long periods of time, or the repetitive motion of raising your arm up to touch the screen. That's not something most deskjockeys are going to be doing a lot. It's horrible for ergonomics.

A standalone touch pad doesn't have that problem.

Most phones are held in the hands with lowered arms, hence it's not a problem for those devices.

Hell, laptops were being sold with touch pads as the primary pointing interface. Not much different from a desktop, really.

I don't think any particular feature of touch pads was the perceived problem. But then, you seem prejudiced against Jobs, so my reply is likely pointless.

They exist

[guruevi]

They just won't run MS Office which is the biggest problem for most office workers. They are currently indeed in developer and embedded stage. The problem is that occasionally you want a little more horsepower (even if it's just to play Flash games) so they buy a 'normal' computer. Also there is no real support available and very little experience by your average sysadmin.

Once somebody starts doing it, the ball will get rolling. Even $200 is not bad but once Raspberry Pi runs a browser and e-mail, SSH, VNC, X and OpenOffice and basically plugs into a display without too much trouble (or is embedded into a display even better) I will be deploying them in our shared computer spaces because that's all they're for - connect to the cluster to run your jobs, check your e-mail and Facebook while you're waiting, occasionally copy something from or to a USB stick. All home directories are already on the network (NFS) so I don't really need much storage.

Re:They exist

[jon3k]

They just won't run MS Office which is the biggest problem for most office workers.

Actually, they can..

Re:They exist

[Billly Gates]

Actually MS is releasing an ARM version of Office and Visual Studio when Windows 8 comes out.

Add to the fact that it can't run some games will be considered a benefit to those in I.T.

Arms have the horsepower with good video chipsets and drivers. PC drivers are 1,000 times slower due to terrible drivers than arms with good GPUs according to John Carmack in regards to texture loading and smoothness. Hopefully this will change and ARM netbooks can run 1080p video fine with good gpu integration which x86 is just now doing with Llamo and Bulldozer and the upcoming Intel chips.

Most people need the I/o and video more than raw FPU and integer performance.

Re:They exist

[Anonymous Coward]

It supposedly runs Iceweasel, which implies X, and proves it can handle LibreOffice. SSH is trivial. I've never used VNC, but I can't imagine it's worse than Firefox for resource usage.

just my 2-cents

[nerdyalien]

ARM is relatively new to the pipeline and super-scalar architectures, and yet to see "out-of-order" execution. Assume it will take some more years to produce a highly efficient processor.

My guess why ARM didn't reach desktop yet, it can't handle complex multi tasking stuff. For an example, how about viewing 20 web pages, HD video playback, downloading multi-gig file at 1 Mbps, applying PS touches on batch mode, while compressing 1000+ files. My other guess is, ARM is yet to produce a CPU that can communicate with modern graphics cards, SSDs and other demanding I/Os. They require additional transistors/processor to get around it I guess.

Re:just my 2-cents

[RightSaidFred99]

x86 will eat into ARM's mobile (and below) market well before ARM reaches the upper echelons of x86's market.

iMX515 - only 512mb RAM

[lkcl]

you need to look at those a little closer. 1) the iMX515 has a hard limit of 512mb RAM 2) if you've never used a 10,1in 1024x600 LCD, you are in for a bit of a shock.

but yes, the efika mx is getting an upgrade - soon - to the 1ghz iMX53. and, also, i think genesi have been doing "dogfood eating" and have found, just as i told them it would be, that 1024x600 LCDs are completely unusable. their developer, matt, treated my ideas like shit (i was approaching them to see if they'd like to collaborate on a faster, more flexible product). he told me there was no market for high-end ARM-based laptops. and then "oh look!" surpriise, the next version of the efika laptop will have a 1280x768 LCD as standard. hmm... :)

Re:iMX515 - only 512mb RAM

[Anonymous Coward]

Go over Matt Sealey's head. No one in the world takes him seriously, he's just an arrogant asshole with a clever mouth.

He is quite literally an English, mail-order gay bride who, once in the US, got a job at his 'husband/owner's' workplace recycling PCs.
That's the limit of his 'developer' qualifications. Bizarre but true...

Fine, I'll say it...

[acak]

... 2012 will be the year of ARM in the desktop!

Re:Fine, I'll say it...

[Eunuchswear]

... 2012 will be the year of ARM in the desktop!

No, that was 1987.

This article is amazingly il-informed.

Re:The OP is talking about red apples, not greenap

[SendBot]

In a lot of ways, a touchpad is just a mouse by any other name. What makes them interesting are more recent developments that allow these "touch" conventions, for instance two-finger scrolling (which I *love*). I never suggested a poor ergonomic setup, nor would I. With a kinect-like thing, a user could just hover their hands over the keyboard and have little transparent hand avatars on the screen. It's the concept that's important, and debating flimsy hypothetical implementations completely misses the point.

I think Jobs is okay. Heck, I even like the guy. But I read between the lines and take what he says with a block of salt. Remember when iPod competitors started having video playback? He played a scene from Raiders of the Lost Ark to poke fun at them and say they were going to the wrong place. Now how many current iPods play video? All of them except the screenless shuffle I think?

I don't recall him saying that touch isn't a viable input method (and no one is providing any links here), but I'd believe that he'd say something like that only to be later contradicted by his own products, as evidenced by what I've quoted earlier.

Re:The OP is talking about red apples, not greenap

[Sancho]

I think this is probably what the original poster was referring to.

http://www.businessinsider.com/steve-jobs-touch-screen-mac-2010-10

If so, then the OP mangled it a little as well.

Not scaling to threads

[tepples]

Why would you not want desktop computing power at 1/10 of the power consumption?

First, because the existing application you want to use depends on clock rate and instructions per clock and doesn't scale to 8+ threads. Second, because the existing application you want to use hasn't been recompiled.

Re:There is a smartbook available (although not go

[lkcl]

unfortunately, this system illustrates why 1024x600 LCDs are undesirable. as does the genesi laptop with the same sized screen. other than the forced-installation of android, total non-upgradeability, inability to have 1gb of RAM and complete lack of interface for putting in an SATA SSD, the AC100 is actually very good. ok, in case you hadn't noticed, that was supposed to be ironic.

cost and integration, not just power

[YesIAmAScript]

ARM licenses IP. Intel sells chips.

If you license a core from ARM you can put it down on a chip, then put down your other logic (north/south bridge, interface logic like USB) on the same chip. Then you can end up with your entire system on a chip.

With Intel you have to buy a CPU, buy a north/southbridge. If you want custom interfaces beyond that, that's more chips too.

So the net effect is that the Intel-based system uses more chips and that means it costs more, uses more power and is larger. Using more power means you need to put in a larger power supply, that costs more. If it's battery-powered, that means it needs a larger battery, that costs more. Larger in and of itself makes something more expensive to make as it requires more materials. And then it being larger means it costs more to ship from where it is made to the customer. And then finally every increase in costs also means more increase in on-the-shelf price because you not only have to cover the higher costs, but the OEM and retail margins on the costs.

The next effect is that ARM devices will be cheaper to buy and to run. And in the case of portable devices, more sleek too.

This may not matter to some customers but to other customers lower costs means a lot.

Performance is an issue. We have ARMs already in the pipe (dual-core ARM A15) which have sufficient power for most uses and ARM will certainly have even faster cores later.

I see a strong future for ARM in laptops and in home computers. No, not in tower computers but those make up a shrinking part of the market already.

Finally, as others have said, be careful about agreeing with Steve Jobs. He's a consummate liar. Just because he says he doesn't like touch for the desktop doesn't necessarily mean much. It means Apple doesn't deliver touch on the desktop today, but it doesn't necessarily mean anything more. Apple could flip on this at any time like on the video iPod.

Re:cost and integration, not just power

[blahplusplus]

"The next effect is that ARM devices will be cheaper to buy and to run."

I don't believe this at all since modern video cards use ARM already and they are extremely power hungry. Speed costs you in power no matter which way you look at it.

Re:cost and integration, not just power

[Required Snark]

You are right. The forces that will lead to ARM powered laptop (and desk tops) systems lie outside the traditional PC platform. A key enabling technology is System on a Chip, which is normal for smart phone ARM platforms. SOC means that a big player who wants to deploy an ARM solution can save considerable integration cost, as you described in your post. This is not theoretical, it's here: the iPad

It's both pathetic and typical how Slashdot Pundits take the narrow view of the evolution of technology. Once they decide that something is "The True Way", they refuse to consider that familiar technology has flaws allowing competing products to overtake existing solutions.

The mass amnesia about ARM chips in the iPad is a classic case of how deeply most Slashdot members cling to common perceived wisdom. They are not innovators at heart, but tend to be much more comfortable with what they already know.

I'd also like to point out that ARM for laptops does not automatically have to come from Western economies. If some indigenous manufacturer in China decides that they can sell a large number of systems in China by taking advantage of ARM integration, they will. Compatibility with legacy code could easily be discarded if the price difference makes the hardware cheap enough.

Ignoring these kinds of possibilities is another example of how the majority of Slashdot readers are narrow minded and egocentric. Fortunately for everyone else, Slashdot counts for very little in the big picture.

Re:cost and integration, not just power

[Ben Hutchings]

With Intel you have to buy a CPU, buy a north/southbridge. If you want custom interfaces beyond that, that's more chips too.

Not so. Intel has SoCs now. Apparently they're quite popular for use in IVI applications.

Re:cost and integration, not just power

[smash]

The OP was referring to the ability to build your OWN SOC. Not use something intel has made for you. If i want to take an intel CPU and include my own custom processing device on the same die, its not going to happen. With ARM, i can do this.

cortex A9 has out of order execution

[Anonymous Coward]

cortex A9 has out of order execution.

Re:There is a smartbook available (although not go

[ChunderDownunder]

Pricing will dictate whether ARM kills the Atom market. Here in Iberia I see lots of crummy atom netbooks on sale for ~ â200-300 with a puny 1GB of RAM. At least at FNAC and El Corte Ingles, the price of a Transformer is â599

A next-gen Transformer with expandable RAM (for booting Windows 8 or Linux) and a quad core A15 would suffice for many needs and doubles as an Android tablet.

If few people still buy pcs

[tepples]

I believe 10 or 15 years from now, it will be somewhat unusual to have what you would consider a full powered desktop PC at home or at the office, and so will doing evelopment work for these no longer very common devices be.

That's a problem. Doing development work currently requires a desktop- or laptop-class computer because Apple doesn't want compilers, interpreters, and such running on iOS. If medium-duty creative work requires a pc,* but few people still own a home pc after some point, then the economies of scale will drop for pc hardware, and a lot of people will be shut out of doing medium-duty creative work at home.

* lowercase for a reason

Sylvania SYNET07526

[Anonymous Coward]

Fits on the desktop great!

shiva plug systems

[Anonymous Coward]

I just got a dream plug 2. It is a Shiva plug Ubuntu based ARM powered server which has HDMI, Ethernet, USBs, RGB and a few other connectors. It can be a full desktop, but I am having a bit if difficulty finding software to do the nieche things I want it to do.

Far more x86-only than POWER-only

[tepples]

The people who run tons of software that is x86 only and has no comparable ARM version?

Like the people who run tons of software that is (sparc|alpha|hppa|power|mips) only and has no comparable x86 version?

Your analogy falls down on scale. Apart from video games that run on locked-down consoles, far more software used by home and small business users is x86-only than POWER-only, MIPS-only, or SPARC-only. This is especially true since Apple's transition to Intel CPUs in Mac computers pulled POWER out of home and small business. Nowadays, it's either x86-only (software for desktop PCs) or ARM-only (software for smartphones and home tablets), or it runs in a JIT-compiling virtual machine (Java, .NET, JavaScript).

Re:Far more x86-only than POWER-only

[smash]

I think you are over-estimating the *average* home user's use.

Average user checks email, facbook, browses the internet and watches a few movies.

This can be done on an iphone/ipad quite easily. Hell, you could even do it on an appleTV, given the ability to pair decent bluetooth interface devices to it, and a web browser.

David Braben on the history of ARM

[BitterKraut]

BAFTA video: http://bcove.me/tux4wa2x The part most pertaining to the current thread is at 14:32

What's an N900

[tepples]

Being the last to learn these things might have something to do with the fact that United States residents tend to be the last to learn that things like this Nokia product even exist. I guess this is due mostly to the stranglehold that AT&T has on the American GSM telephony market.

Koomey's law of Energy Effeciency

[__aazsst3756]

Arm looks attractive for its low power consumption, but Koomey's law show's that it is energy efficiency, not power that doubles every 18 months.

http://www.technologyreview.com/computing/38548/?p1=A3

Intel is selling chips to build non-Arm (refrained from using "real") desktop computers, with low enough power the savings are simply not worth the tradeoff for most people.

Power is the issue

[SmallFurryCreature]

If you ignore the rampant fanboys then ARM is just slower then x86 and will likely remain so. Meanwhile desktop application developers like the power.

I have recently switched to a relatively lowpower E350 and while the power drain is far lower then a dualcore regular desktop, you can notice it struggling with some stuff. Take for instance a torrent client like Vuze. Bad example because it is a piece of bloatware? True enough but really capable torrent clients are hard to come by on Linux. Take uTorrent and its absolute bandwidth hogging traits (ran it for a while but the entire connection would slow to a crawl on even the simplest jobs. KTorrent opens a move window for every single file in a torrent it wants to move, turning you desktop into a flashing nightmare and showing KDE just plain ain't ready for the desktop.

But crap or not, almost all make the little desktop struggle at times. Not that it doesn't run but as my main desktop, it would get annoying REALLY quickly. So my main desktop is a massive 4 core machine running mostly on idle...

It sounds counter intuitive but a desktop runs best for the user when it is barely doing anything at all. Many people who think ARM could take over seem to reason that since they desktop is using only 1% of the resources then a computer with 1% of the computing power would give the same smooth experience.

It would not. If your PC showed the same performance as your phone or your tablet you would go stark raving mad.

ARM will take over on the desktop when people will buy cars that have just enough horse power to do realistic city acceleration. Go ahead, buy a car with a 10 horsepower engine... when you think that is okay, you are ready for an ARM PC and the retirement home.

Re:Power is the issue

[Anonymous Coward]

Large corporation like the one I'm working for use a x86 PC only to login at a Windows 2008 terminalserver. This could easily be done with a ARM desktop if only produced in large quantities to make them really cheaper. It will not happen because our Wintel monopoly will not allow it.

Archimedes

[wavedeform]

The Acorn Archimedes: http://en.wikipedia.org/wiki/Acorn_Archimedes This was some sort of outgrowth of the BBC Micro - http://en.wikipedia.org/wiki/BBC_Micro

Why do you need a desktop?

[Anonymous Coward]

I think it's better to get rid of the desktop, as long as you have a socket to plug your computer: http://www.plugcomputer.org/

Brilliant post - the math works

[Weaselmancer]

I have no mod points at the moment but yeah, brilliant post. That's it exactly.

I'm going to be buying a beagleboard here soon to start the household transformation. I have a x86 server I run in my basement that I'm going to replace. It will instead be a beagleboard. It'll drop from about 100w to maybe 6w or so.

Here's the math on the savings:

.1kw * .09 dollars/kwh * (24 * 365) hours = $78.84 dollars to run my x86 server for a year.

.006kw * .09 dollars/kwh * (24 * 365) hours = $4.73 dollars to run my ARM server for a year

So switching to ARM saves me $74.11 every year. A beagleboard XM costs about $150 at Digikey. So it pays for itself in two years, then saves me about 75 bucks a year every year.

Re:Brilliant post - the math works

[rcs1000]

I'm sorry - your numbers are *way* high for the Intel server. If you use a Mac Mini, headless, then you'll be at 0.01kw/h when idle, peaking at 0.085kw/h at max. Given any regular server will spend 90%+ of its time idle, then your true Intel number is probably $10-15.

Any server you build yourself will have similar power characteristics to the Mac Mini.

You should still go for ARM, because it's more fun, but don't fool yourself into thinking you'll be saving money.

Re:Brilliant post - the math works

[Weaselmancer]

Dude if I had a Mac Mini laying about I'd certainly be using that for my basement server. But I'm not. I'm using a crappy P3 550MHz eMachine. That has grinding fans and hard drives you can hear from across the room. I'd imagine it pulls as much juice as a 100W light bulb at least.

But you're right - I should slap a meter on that puppy and get some real numbers before getting happy about my savings.

Re:Brilliant post - the math works

[yakovlev]

A server you build yourself is NOT going to have similar power requirements to a mac mini. Quite frankly, 10w at idle is INSANELY low. You'll be lucky to be as low as 20w idle for just the processor and motherboard on a new build.

I spent $200 and upgraded my desktop from a 140w behemoth to a new machine that runs at 55w at idle, 40w at standby, including hard drives, fans, etc. I figure I will make the money back in 4 years, but the much nicer machine makes it worth it already.

For something like a server running on a Beagleboard, the hard drives will probably wind up being the main power hogs, and you might be surprised how much they draw.

pandaboard

[aaronpeacock]

pandaboard.org brings hobbyists (not for OEM use apparently. you have to talk to TI about your own OMAP implementation if you want to OEM) a wicked ready-to-roll platform... load linux and go. fanless and high performance.

pcie interfaces

[Anonymous Coward]

marvell alone has >5 chips that have 1-4x pcie.
(i'm currently using 78xxx series and sheeva series with
pcie interfaces.)

could you please elaborate on how you're counting?

Re:pcie interfaces

[lkcl]

marvell == 1 :) sorry, i keep forgetting about marvell because of their NDA policy. when they quit with that shit, i'll be able to count their CPUs, because i will have seen the f*****g data sheets! i knew they had at least one, though.

x86 and ARM in same box would interest me

[SoopahMan]

An ARM-only desktop seems silly, but I'd be interested in an ARM/x86 system that can use both with a single OS. With Windows 8 supporting both it would be cool to run mobile and desktop apps natively on the same machine. It would likely ease the transition away from x86, ultimately. But the blurred line would really give consumers a lot of options. It might also be interesting to see such a device running Ubuntu and supporting Android apps.

Re:x86 and ARM in same box would interest me

[sourcerror]

My guess is that the Win 8 you run on desktop won't be the same that you'll run on ARM. So you have to take the lowest common denominator, and that'll be stuff that run on Metro. My guess is that your use-case isn't popular enough to create a distinct Win 8 version.

Desktop? Where are the servers?

[erice]

Last year, ARM servers were the big thing coming. There were stories of custom made servers. Marvell announced and later demonstrated a multi-core, 2+ Ghz ARM chip for servers. ARM servers were suppose to be available for the rest of us "next year". Well, it is "next year". Where are they?

ARM was developed as a desktop CPU

[Alioth]

ARM was originally developed as a desktop CPU, and it was on the desktop - it's been and gone.

ARM stood originally for Acorn Risc Machine, it was developed by Acorn because they couldn't find an adequate processor for what they wanted to do to follow on from the 6502. Many of the CISC chips at the time (mid 1980s) had very poor utilization of memory bandwidth and poor interrupt response (Steve Furber in one of his talks recently on the development of the ARM - he's one of the two people who developed the first ARM CPU, pointed out in particular the National Semi 32016 (IIRC) that they were thinking of using, until they found out the multiply instruction took over 100 clock cycles and could not be interrupted).

They also wanted ARM to be low power, not to make their new line of desktop computers energy efficient particularly, but because they needed it to be cheap so the computers could be affordable. If they could get it under 1 watt, they could use plastic packaging instead of ceramic packaging which reduces the cost by an order of magnitude. They had no tools for estimating power, so they just designed *everything* on the chip for low power. When they got the first samples back from the fab, they were blown away when they found the chip consumed 0.1 watts - they had massively overachieved.

We had the Acorn Archimedes in school. IIRC, it had an 8MHz ARM and it could emulate - in software - an IBM PC with VGA graphics faster than the original IBM PC ran. That's how much faster the ARM was at the time compared to anything else around. Without needing to be in a ceramic package.

Re:Desktop? Where are the servers?

[Eunuchswear]

Last year, ARM servers were the big thing coming. There were stories of custom made servers. Marvell announced and later demonstrated a multi-core, 2+ Ghz ARM chip for servers. ARM servers were suppose to be available for the rest of us "next year". Well, it is "next year". Where are they?

Mine is in the garage.

Sheevaplug with 1To raid1

Ah, memories...

[Aardpig]

...I was the first person in the world to have Linux running on an ARM 250 (that being the chip inside my long-gone Acorn Archimedes A3020 desktop). This was in the very early days of ARM Linux, and I was finding that the kernel would invariably oops when I booted up. Rather than submitting the problem to a forum (there weren't any), I caught a train to somewhere in south London, with the A3020 in my backpack, and spent an evening in the bedroom of Russell King -- the guy who originally ported Linux to the ARM architecture. Took him about 30 minutes to figure out what was wrong and get the kernel to boot.

Re:There is a smartbook available (although not go

[shutdown -p now]

Transformer is awesome, but Android is not a solid desktop OS for use in docked mode (yet) - just showing the mouse cursor does not make it so.

And sure, you can run Ubuntu in chroot, but you have to use VNC for GUI, and it's terribly slow and inconvenient. There's also dual-boot, if you own a rootable Transformer model, but last I checked they were still ironing things out there.

Yes

[Billly Gates]

For x86 to die, 2 things need to happen
1. Apps apps apps
2. Needs to be better and cheaper and offer abilities that consumers would be willing to drop x86.

The reason others failed is because they were niche and expensive with no apps. Alpha was fucking awesome 6 years ago! But it was for advanced workstations and servers. PowerPC was awesome too. It had more success in macs and gaming consoles as it got close to fulling being the better CPU for cheaper a little and had niche apps for gaming in Nintendo boxes and running mac apps. It didn't have enough price and performance to sway x86 users away nor the apps. The pentium Pro with NT 4 prounced it and offered x86 compatiblity. It was good enough

Arm however has the apps and the price point plus power management capabilities. ARM is winning by going around the desktop and targeting phones and other devies with hundreds of thousands of apps. It has the apps.

If ARM grows with Windows 8 and MS offers visual studio to compile apps for it it will satisfy both conditions of having apps available and being able to offer what x86 can not. It will take several years. My guess is not Office but rather Windows Server will have ARM versions and enterprise apps will be ported. If Oracle gets Java to run on the ARM more Enterprise apps will be ported. Games like Angry Birds will come next on win64 and the rest will be history. More apps are run as html5 and ajax anyway. Fat binaries can help like it did with Apple. Within 2 years most Mac apps could be switched over without a problem.

Maybe in 5 years Apple might consider ARM versions of MacOSX as it tries to merge it with IOS too.

low watt computing +1

[itzdandy]

I have switched to lower power computing. I once ran massive computers that were honestly, way overkill for my needs.

I now run atom and zacate based systems. In fact, storage uses more electricity in my house than cpus. This is my next project to tackle as I switch to lower rpm, 2.5" and ssd disks.

I'm not a tree hugger, I don't play Al Gore scare tactics to convince people that the world is ending blah blah blah. what I am doing is measuring my short term cost vs long term electrical savings. I don't but low power just because and I don't drive a prius because they are not less expensive to drive (until the 2012 pluggable prius, 45 miles on electric is about cost parity with my gas car)

I do have a very large media library. I have many TB of ripped content from boxes of dvd videos now happily stored in boxes in the garage. Lots of content pulled from my DVR, web downloads etc and all exposed to my TVs and computers via appletvs cracked for XBMC which is a very low power ARM platform, or the zacate based fileserver/player.

Wrap this all up and switching from my 25W atom/zacate fileservers (3 of them) to ARM setups can save me about $50 per year. Additionally, These file servers themselves can output to the TV eliminating the need for a few appletvs. Right now my zacate fileserver is the only one able to output to TV, the atom's dont have to video capability to playback my videos without added hardware that increases the expense and power consumption.

Then again, most people could realize $50-$100 savings by turning off power strips and cutting out a lot of standby waste.

Re:There is a smartbook available (although not go

[fnj]

Yep, from the review, it's crap plain and simple.

Um, what?

[Anonymous Coward]

I guess those hundreds of picoITX ARM based desktops I have to maintain at work are just a figment of my imagination?

Mobile phones will be your desktop

[Anonymous Coward]

I give it 5 years before the nearly every mobile phone comes with a docking station that features:
* HDMI
* USB
Some will have internal storage, and I suspect that something like lightpeak will enable the docking station to have extra processors.

Fact is, my Samsung Galaxy SII is more powerful and has more memory and disk space than the PC I started the century with.
It is more than adequate for my basic desktop needs. I dare say that the form factor has driven efficient coding that makes it even more powerful.

It's hard to figure out who will drive this.
Apple could do it tomorrow, but they have a vested interest in protecting their mac sales. Without Steve though, it's anyone's guess. Will the new guy slaughter the low end mac and desktop market simply to get there first? I doubt it.
Which leaves the Android crowd, or possibly Microsoft. Neither of whom sell desktops.

Desktop OS in MS' case, but with the design forces behind Windows 8, perhaps this is their new killer app?

Sorry, but you are wrong about ARM and cost

[tlambert]

Sorry, but you are wrong about ARM and cost

The following is my personal experience:

The problems with wide-scale adoption of ARM for anything but higher cost or full SoC embedded systems (like the baseband radios on a cell phone) are:

(1) Economies of scale have a high impact on x86 architecture based systems costing less

(2) For comparable cost CPUs, the performance of ARM is less than Intel (see #1)

(3) Almost all ARM based designs end up with more discrete parts for comparable functionality x86 designs, due to lack of an ISA for ARM leading to something similar to the x86 ISA's common bridge chipset functionality

(4) Per The Innovator's Dilemma: peripherals continue to march up-market

That last one is really the kicker: what did a bottom-end laptop cost in 2001? 2006? What does it cost in 2011? It's the same inflation-adjusted cost inflection point; only now, the CPU is faster, there's more RAM, and more storage. You simply can't buy 16M of DRAM these days or a 1G disk, unless you are willing to pay _more_ than you would pay to get a higher amount of resources at the lowest saddle point in the COGS.

It doesn't MATTER if you need less resources, the cheapest commodity price at which you can get a particular TYPE of resource is almost never achieved by getting the smallest capacity. You can't take advantage of decreased cost per resource unit, because you simply can't obtain the same number of units as you did 10 years ago.

You can try to design SoC's to get around the bridging problem, although a quick look at the Linux kernel source tree to inventory lines of code dedicated to system-specific x86 ISA features vs. board level features will quickly disabuse you of the notion that cost will be driven lower by taping out new ARM-based SoCs to try to achieve integration. There are a bazillion "board this" and "board that" files in there to handle random gpio and i2C/SMBus/PMBus/IPMB/we-are-more-cleve-so-we-invented-our-own/etc. peripherals.

This is without even getting into other issues, like the graphics engines being optimized for different operations than most Intel graphics people are comfortable with, or companies that want to run TZones to effectively get a hypervisor-like control so that can reduce cost by running baseband code on the same CPU, and you have to trust them not to put in back doors and not make coding errors which could result in you being vulnerable because they're vulnerable.

The bottom line is that, at the present time, comparable ARM hardware to x86 hardware generally results in similar COGS, even without the so-called "Intel tax", and throwing out everything you think is not necessary to try and reduce costs just gets you down to about the same COGS as an x86 solution.

This might not always be the case; maybe the ARM vendors will call off the SoC-defacto-standard-vs-ISA war they are currently engaged in, and give in, in much the same way that there are no longer 120 versions of UNIX, like there were in the late 1980's/early 1990's. However, I would not hold my breath, as it's not showing signs of happening any time soon.

-- Terry

Re:Sorry, but you are wrong about ARM and cost

[Megane]

(3) Almost all ARM based designs end up with more discrete parts for comparable functionality x86 designs, due to lack of an ISA for ARM leading to something similar to the x86 ISA's common bridge chipset functionality

You keep using that word. I don't think you know what that word means.

Are you joking ?

[Kuruk]

Locked to Microsoft's app store. Good joke.

The app store still is not working.

Re:Has Everyone Already Forgotten About the Netwin

[The_Dougster]

I still have a Netwinder. I spent quite a bit of effort porting the system to Gentoo ARM softfloat, although at the end of the day, it really hasn't aged well. The 275MHz ARMv4l seems a bit pokey now but it can run a bare-bones xorg system with a lightweight window manager. Prior to my porting effort, you had to use hard-float glibc which would trigger an interrupt and the kernel fpu emulator would handle the call. I submitted some patches upstream vs gcc but I have no idea if anybody ever did anything with them, for all I know I have the only softfloat-linux Netwinder in existence.

Just last year

[Anonymous Coward]

Just last year around this time there were several desktop ARM's for sale at C-----SA with Windows 7. Just like netbooks, except needs monitor and keyboard and mouse. They make great terminals for Virtual Desktop access.

mod parent up!

[reiisi]

This is what I was going to say.

Tower case with slots for lots of CPUs.

Or a lunch-box or double-lunchbox tower case with slots for five or ten CPUs, and I can put my firewall/modem/bridge on one CPU, my outward facing httpd server on another, my inward facing DNS server on another, my local file server on another (takes extra room for a terabyte 2 1/3" drive), local timestamp server on another, and one more for a watchdog to watch them all. Optional filter/proxy server. Maybe even put the family mail-homework-corkboard server (with video output for a digital frame screensaver, etc.) in the same box.

Before you burt yourself, ...

[reiisi]

Learn how to read and use benchmarks.

Sure linear algebra will give you an indication of how fast you can expect certain games (synthesized graphics) to run, but it doesn't say much about what on ordinary non-gamer can expect.

Two cores are going to be more responsive for interactive systems than a single core. You have to go way more than double the effective core speed to get single core CPUs to respond as quickly to user input as you can get dual core CPUs to respond. Linpack has nothing to do with that.

On the other hand, two cores do not double CPU bound compute in general. With a properly compiled linpack on a well-designed CPU, you can saturate the compute on each CPU, and thus pretty much scale the speed with the number of CPUs, which, as another poster points out, causes one to question your choice of numbers to compare.

Ergo, your expectation that you can just compare linpack to linpack is misguided at best. To analyze the results properly, the compiles have to be co-ordinated. The compiler switches for the one have to be matched by the compiler switches for the other, and have to be selected for the expected problem set to be meaningful. And then they are meaningful for the selected problem set and may not be meaningful for other problem sets.

Sure, we can say that the K is significantly faster than the Tianhe, but we can't say it's 4 times as fast in any particular real application. We expect, in fact, that there will be some applications where the Tianhe is faster.

Linpack is not what the average user is running. Shoot. I have an old clamshell PowerPC G3 ibook that takes a long time to start Firefox, but once it's running, it's reasonably responsive. Bound more by my ancient ADSL connection than by the 300 MHz CPU and 67 MHz RAM. (The thing is slow installing Linux, however.)

SoC pitfalls

[unixisc]

System on a Chip has been tried before - Cyrix, DEC AXP 21066 (or was it 21068?), Centaur... Problem w/ this approach is that the CPU has a much watered down performance since it now operates @ the same level as its support interface logic. SoC is good for embedded systems, where performance is not a major issue - cost overrides it in a big way. Even power savings won't be there, since whatever extra circuitry for performance is being eliminated gets replaced by the interface circuitry, so that support chipsets ain't required.

It's a good question

[LostMyBeaver]

More importantly, what's the real purpose of a desktop computer anymore. In my house, I have 14 PCs at the moment. 6 media center PCs, 4 laptops and 4 miscellaneous.

Media center PCs are a great place for ARM chips. A media center PC doesn't require a great deal of CPU capacity, but it does require a good video decoder and most ARMs these days are shipping with pretty good hardware media decoders.

Development is another obvious reason. I need something to compile ARM core on for people running on ARMs. Given the current ARM offerings, none of them are fast enough to actually develop on. I'm currently using a Core i7 Sandy Bridge with 16 gigs of RAM and SSD. It's less sluggish than my last i7, but frankly, I am considering a desktop with either two or four 6 core processors in it. 5-10 seconds shaved off each compile adds up to hundreds of saved hours. I save and compile CONSTANTLY. Matlab is also much faster as you through more CPU and GPU at it.

Personal servers are another great reason. A mini-itx motherboard with 8 Serial ATA controller ports on it would be very useful. This should in theory allow building a nice little 22 terabyte RAID with very little power consumption using 8 3-gig drives, a USB thumb drive to boot from etc... with staggered spin up, a 22 TB RAID 5 might be able to function on a 50 watt power brick.

Disposable computing would be another great reason. Using pico or nano-itx motherboard based systems with ARM processors would be nice since you could have tiny desktop machines with everything you need in a small box. Every time I bring furniture to the dump, there's a bin large enough to park 3 cars in rapidly filling with computer equipment... and that's purely residential... most of the systems being thrown away are being thrown away because of one bad part (I suspect, most often a blown capacitor) or because they're too slow or just not fashionable enough anymore. With the much smaller boxes, a great deal less PCs )measured in cubic meters, not number of PCs) will end up in the bin after what I call the LCD effect passes... you know, when the world threw away a gazillion CRT screens that still worked because they're not as fancy as LCD and plasma. People will still toss machines by the bazillions, but they'll be much smaller. If the companies making these machines would also commit to alternatives to epoxy for PCBs which can be broken down easily, then it could be a wonderful step forward towards green computing.

The issue coming up isn't about ARM vs. x86. Personally, in heavy computing work loads when the processor is running at 100%, the power meter on my PC (I have one on each PC) runs at about the same number of joules per function. The ARM works great in low power requirements, but when running heavy loads, it's not that fantastic. If someone ever releases a decent ARM compiler, this might change.

What makes the ARM ideal is the configurability. The ARM market is more about jamming everything on to one chip, not providing I/O for connecting other things. So, you can often find just the right ARM for the job... but if you can't, then you'll have to wait until you can... this might change if the new ARM chips contain PCIe controllers with many lanes, but for now... just imagine the next Mac Mini being in the same box as the Apple TV ships today. There's no reason it shouldn't be. Add some more RAM and some more flash. Back in 2000, I saw Mac OS X running on x86, PPC and Sparc... I'm sure that since Apple already has all the compilers and developer tools in place, they have an ARM version too. A Mac Micro (instead of mini) could be an ARM based Mac with 64 gigs of Flash and 2 gigs of RAM with an A5 or A6 processor. The Mac Mini already could do nearly all of this anyway.

So... there are some reasons and some options.

BTW.. ARM performance in my testing is about the same as x86 if you use the same amount of power... for general purpose tasks... in heavy loads, it still doesn't touch Intel or AMD.

ISA=Industry Standard Architecture

[tlambert]

ISA=Industry Standard Architecture

http://en.wikipedia.org/wiki/Industry_Standard_Architecture

-- Terry

Re:Desktop? Where are the servers?

[DuckDodgers]

Marvell said their Armada XP chip would be out "next year" but 2011 is not over yet. nVidia never promised that their desktop and server ARM chips would hit market in 2011. I don't think they provided a release date at all.