Just – just hold on a second. This is big: NVIDIA, maker of graphics accelerator chips, has just announced, during its keynote at CES, that it is developing a high-performance ARM-based processor together with ARM, targeted squarely at the desktop, server, and even high-performance computing markets. That Windows on ARM thing? NVIDIA referenced it multiple times! Update: Boom, and we have a press release. “NVIDIA announced today that it plans to build high-performance ARM based CPU cores, designed to support future products ranging from personal computers and servers to workstations and supercomputers. Known under the internal codename ‘Project Denver’, this initiative features an NVIDIA CPU running the ARM instruction set, which will be fully integrated on the same chip as the NVIDIA GPU.”
The keynote only just ended, so we’ll probably be updating this article as events unfold. The news is this: inside NVIDIA, a team has been working in collaboration with ARM to develop a high-performance targeted at the desktop, server, and even high-performance computing. The project is called Denver.
It’s in development at this point, so this is all I could gather from the liveblogs. Stay tuned for more – I’m sure Microsoft’s keynote tonight has something to say about this as well!
Exciting times!
Sounds like a powerful combination, but it will probably cost a leg.
I really don’t mean to rain on anyone’s enthusiasm, but why are you excited about Windows on ARM?
I don’t really see how it can be any kind of game changer or even of academic interest. In fact, arguably the real question is why it took Microsoft so long to get Windows running on ARM. Linux has been running on ARM, well, comparatively forever. Darwin (Mac OS X) has been running on ARM on some form or another since the first iphone…
Are you thinking that we’ll see Windows 7 running on ARM notebooks? Perhaps… but the desktop/laptop/notebook era is over. Seeing Windows on ARM now is more of a “Phhttt”/shrug shoulder while thinking of the next cool Android/iOS (heck, even Blackberry QNX) app.
Sorry — I just really don’t see how this is at all exciting.
We don’t know how long Windows has been running on ARM. Considering the excellent portability record of Windows NT, I’d say it’s been running on ARM pretty much since day one – even if only as a technical exercise.
Microsoft’s first active foray into breaking up WinTel? You don’t see how that is exciting?
Granted, NT 4.0 was the last official release to be officially multi-platform (Alpha, MIPS, PowerPC), but that was back in 2000. Supposedly NT 5 (Windows 2000) had an RC2 build that still ran on Alpha…
Needless to say, the fact that NT5RC2 was nominally cross platform does not imply at all that current Windows 7 (NT6.2) has maintained that legacy. Although it would be bizarre if it hadn’t, but perhaps not surprising.
MS breaking up WinTel doesn’t really make sense. ARM processors are supposedly ideal for mobile use, and MS has Windows Phone 7, right? Recompiling Windows 7 for ARM to squeeze on very low end “PCs” doesn’t sound like a compelling business model. Especially since most of the target user have no comprehension that different core CPU architectures mean differently compiled software packages. “What do you mean I can’t run Office 2010 on my *Windows* notebook?!?!”
I see the whole NVIDIA announcement much more in the line of future converged devices/interfaces a la Chrome notebooks. Say tablets/phones/notebook-ish devices running Android apps across all. That actually I find much more exciting than imagining Windows on ARM — but hey, that’s me. If running Office is exciting, each to his own… 🙂
Edited 2011-01-05 22:11 UTC
The Windows Server 2003 and 2008 releases, which are just the same as the desktop versions with different tuning and applications, have been running on Itanium.
That is a very different CPU than x86/x86_64.
The .NET system has been running on Itanium and on Xbox 360 or PPC, so we know that can be ported also.
So from that we know that Windows 7 cannot be too difficult to port. We also know that any applications written in .NET “pure” (no p/invoke) will run fine.
..to quote MS own press release:
” Microsoft Office running natively on ARM also was shown as a demonstration of the full depth and breadth of Windows platform capabilities on ARM architecture.”
..i have to say, the idea of tegra 2(or similar) tablet that runs android or chrome-for-tablet and can dual boot win(8?) sounds pretty appealing and perhaps not that far fetched now
go back and read it again. the new arm stuff is not for low end, bueller.
Actually it make very good business sense. Provide choice to the Consumer. The more platforms that an OS can run on, the more applications we will see that OS find. It is very simple math.
MS is a software business, notwithstanding their forays into hardware construction. They need to compete in as many arenas as possible, if only to hedge their bets. Imagine if a new ARM based server became the best hardware for a particular environment? If there was no Windows option, then they can’t hope to compete.
It makes perfect sense. And they will have to compete with Linux. I don’t think they can take anything for granted any more. Every battle they concede, could be a nail in the coffin.
MS plays catch-up in so many places, they cannot afford to take some chances.
???
Whatever, what are people going to use to create all that content you consume on your phone? Do you really see office workers and the like doing thier jobs on phones?
There will always be desktops, because you can’t do real work on a phone, or even a pad.
I wish I could vote this up 1000 times. So many people assume that because one market is growing faster than the other, that the slower-growing one must be obsolete/dying.
In reality, people will continue to need real computers (i.e. decent sized screen, real keyboard) for the foreseeable future. I can imagine powerful laptops replacing desktops in most cases, but even that is a stretch. Anyone who believes that iPads and smartphones are going to replace real computers is delusional…
Edited 2011-01-05 23:43 UTC
It depends on how do you define desktop.
For instance, if some phone/pad have external screen and keyboard/mouse connection (we can use hdmi and bluetooth for now) you can do “real” work on it. Technically it is possible to work on such device combo with comfort. We often forget about first PCs with 640Kb ram, and 7Mhz processor. Plenty of work was done on those machines.
But will it ever happen. Of course not. I will like to see manufacturer that is willing to sell you one device instead of three. Now you have to buy desktop/laptop for real work, pads/tablets for pleasure time and phones for staying in touch. And don’t forget about software that needs to be written for such device.
So is it possible – yes. Will we live to see such product – no.
Edited 2011-01-06 00:55 UTC
I’m sorry, but if you are working in a major IDE, image editing, video editing, 3D editing, audio editing, what have you, you are going to need more than even 1280×700 to do any real work.
It’s that simple, really, you aren’t going to be lugging around 22 inch tablets, and you aren’t going to want to spend more total on all the accessories you would need to make a normal tablet equal a mediocre desktop
uh you can use a monitor that is separate from the computer. it doesnt matter to the monitor how big the computer is.
that’s one of those accessories I was talking about…
I kind of like this idea of a phone which becomes a ygdesktop computer when you plug it in a dock. And from a performance point of view, it might work for most everyday tasks, even though those phone manufacturers who say that 1GHz and 512 MB of RAM are not enough for rendering things snappily a small phone’s screen make me feel bad about the future of software optimization.
The problem I can see with that is software, though. At the moment, you have desktop UIs, tablet UIs, and phone UIs, which work in a different way. For this to work in a smooth and integrated way, we should have UI toolkits and software which adapt themselves to the hardware they run on, even when it means changing screen size. At the moment, looking at the announcement that Android 3 will be for tablets only, we’re not really going in that direction.
Notion Ink’s “panel” idea (putting several phones applications on a tablet screen and having tablet applications have their own small panel for quick access) is a step in the right direction, on the other hand. But will it be followed ?
Yeah, but by the time you hook up a monitor, mouse, keyboard and whatever else to your tablet, you have something that isnt really much better than a desktop.
Too bad I already posted a comment, can’t mod you up. Well said.
I think Motorola just showed us the future.
http://gizmodo.com/5725505/motorolas-atrix-android-phone-has-two-co…
Phone with a laptop dock. Phones continue to get more powerful. The only draw back is input size and display size. The dock solves both.
The Nokia N8 got HDMI and USB OTG.
So you can hook up a screen, USB-HUB, mouse, keyboard to it. And USB memory stick / SSD / HDD.
Personally that is kinda itching by purchase nerve since it would be a portable desktop =P
But I would had felt more secure with it running MeeGo. So maybe I wait anyway, or I’ll get one used.
…you can’t do real work on a phone, or even a pad…
I find your lack of faith in imagination and definition of, “REAL WORK” disturbing,
As someone who works daily in GIS, I welcome any change in technology that I can imagine helping to do my job more efficiently and effectively.
Also, there is a big difference between a Personal Computer (PC) and a Workstation.
I don’t really think that there is a real difference anymore between PCs and Workstations, considering you can get a 4 core, 8 Gb machine with hardware raid and all that jazz for under a 1000 bucks, and just a few years ago, that was a serious workstation.
I’m not saying there isn’t some real advantages for mobile computing, I’m just saying that desktops aren’t going anywhere for a real long time.
That may sound funny, but its not about Windows at all …
Just think about it, why x86 architecture is popular? Because Windows run it, its widely available and cheap, with a lot of possible configuration market top to bottom, but do You HAVE to run Windows on it? NO!
And that the point, by Windows ‘ported’ to ARM will make that architecture a lot more popular, so hardware will be a lot more popular, so You cound get needed hardware for any operating system You want.
Just think about quad core 1.2GHz ARM laptop that is abloe to work for about 24 hours straight on battery
Excited… More like Surprised… I don’t hear anyone jumping for joy over Windows on ARM, but hear a lot of surprise…
Surprised, because microsoft is languishing in legacy land.
The big story is NVidea building an SOC ARM… Sweet….
As much as the “Gamer” crowd laughs at the thought of using an ARM in a PC or Server, I have been Excited about this development for quite some time…
good to see it being announced MainStream.
Gamer ? More like the performance-savvy crowd as a whole. Heavy image and video processing are not for ARM chips at the moment, though it might become the case in the future.
On the other hand, for a low-powered computer, like a netbook or a cheap desktop/laptop for office work and light entertainment, this could be very nice.
Edited 2011-01-06 18:06 UTC
Especially since Nvidia stated that they have no plans to support anything but Windows on ARM: http://arstechnica.com/gadgets/news/2009/06/nvidia-says-no-to-linux…
Made worse since asking recently on the Nouveau’s (reverse engineered Nvidia GPU driver) that there is nobody working on supporting the Tegra platform.
I hope this restores competition with regards to the x86 / amd64 architecture in the desktop / laptop / netbook market.
Are you kidding me? You can get a dual-core cpu for $50. I spent $70 on drinks, dinner and movie last weekend. A cpu can easily last 10 years if you take care of it.
Heck, longer than that.
I’ve still got a dual Pentium 166 MMX system running on a Tyan mobo from 1998. It’s my internet gateway machine with DHCP, Squid proxy, Apache web server, QoS routing, IPSec tunnels, the works. It’s even running a BitTorrent seeder and tracker.
Yes, so there is competition between x86 / amd64 products. But not so much competition with regards to x86 Vs something else.
I cannot currently go to my local computer shop and see a selection of x86 or ARM laptops for example.
Competition brings keeps prices down and technology improving.
Atleast I doubt Intel and maybe especially AMD is happy with Nvidia getting into the CPU business, or well, they already are with Tegra but even more now.
NVidia have done x86 compatibility before “On August 8, 2008, Transmeta announced that it had licensed its LongRun and low-power chip technologies to Nvidia for a one-time license fee of $25 million.” http://en.wikipedia.org/wiki/Transmeta
Transmeta worked by compiling the x86 CISC instructions into a custom RISC set for the CPU; which is basically what Intel have been doing since the Core Duo CPUs. Thus, it isn’t a stretch to convert x86 CISC into ARM RISC mnemonics.
That is if compatibility is going to be kept. It could be a possibility, remember that the original Itaniums had an x86 mode that was pretty slow, but worked.
Great examples. They show both possiblities.
Transmeta:
Translation was too slow. Company didn’t get many sales.
Intel:
Core 2 Translation was great. Company retook performance crown from AMD.
Either one of those is possible here.
How well did the Transmeta perform to begin with then?
Can’t it be that the Transmeta CPU was too slow and the Core 2 Duos where made fast?
Also is it likely that these ARM chips will have the raw power to compete against high-end AMD64 (that’s what the instruction set is called .. Call it x86-64 if you want to) chips?
I assume not to begin with, and hence no performance crown. Later? Maybe. Regardless of translation or not.
Well, its difficult to say what transmeta’s “native” speed would have been because there was never any native software for it. But, the point I was trying to make is that the translation, even when designed for by the chipset manufacturer is no guarantee for success.
LongRun had nothing at all to do with x86 – it was patented IP that had to do with dynamically adjusting clock frequency/voltage based on load. Nearly anyone that built CPUs/GPUs ended up licensing it outright or trading patents for it, except Intel. Intel ended up getting sued over it and had to pay Transmeta $250 million to make the issue go away (they should have just licensed it like everyone else did).
I don’t want to come across wrong, because I completely agree with that last sentence – it isn’t a stretch and is probably pretty easy to do. But Intel has never made a CPU that functions in the manor that Transmeta used to do Code Morphing.
Transmeta did the conversion in software – it was a VLIW core underneath, but the instruction set the core executed was purposefully designed for efficiently emulating x86 (it had quite a bit of specialized widgets for it), and most of what would be considered the “front end” of an x86 core was done entirely in software. It was in many ways similar to Apple’s Rosetta or DEC’s FX!32 – just much closer to the metal so to speak.
Intel and AMD do nothing like this. Yes, they both internally use RISC like instructions, but the conversion from x86 to this internal instruction representation is done entirely in the front-end/decoder (i.e. in hardware). Transmeta did it their way to save power, Intel and AMD do it for performance reasons – the two approaches are completely different.
But that was mostly a hardware emulator – Itanium basically had a complete x86 frontend in it. The current Itaniums use software emulation, but I use the word “use” very lightly here since almost no one actually bothers to use it (its very very slow).
Anyway, I don’t think emulating x86 on ARM would be technically that hard – BOCHS already does it. Doing it fast is a different story. I personally don’t expect to see any form of emulation for x86 come from MS. A 3rd party might do it, but I suspect Microsoft will just stay away from it.
If anything has been demonstrated over and over throughout the history of computers, its that if you are going to support backwards compatibility through emulation – you better be able to do it at speeds that are close to performance transparent… Apple managed to get close with Rosetta, and it served them well enough during their transition, and DEC did ok with FX!32 – mostly because their CPU was just so damn fast that it made up for the emulation overhead. But the list of failures in this area is a mile long…
MS is better off cutting the apron strings and just forcing software to be recompiled for ARM. In the short term it might hurt for a while, but in the long term it is much better for them to focus their efforts on getting their tools to treat both platforms equally and just let developers target them both at will.
Also, their recent decision to ditch CE binaries on Windows 7 Phone makes me think they have maybe finally started to see the light a bit – backwards compatibility is great until you start to get crushed under the weight of it… This is one thing Apple gets right. Platform transitions (or in this case expansions) should be well planned, precisely executed, and brutally short. The ISVs need to sink or swim – MS does too much hand holding already and they can’t keep it up and expect to compete with the sharks that are in the water now (Apple, Google, etc.)
What can be done in software can be done in hardware and vice-versa.
What I meant by my comment is that if any compatibility is desired at all, there’s a number of different approaches that vary between the amount of software / hardware.
Apple’s Rosetta worked well because the new Intel chips were much faster than the ageing PPC chips. I doubt that these new ARM chips are going to be that much faster than an equivalent Atom chip to supplant the overhead of converting x86 to ARM, it would have to be done at hardware level.
I have to admit that I don’t understand neither NVidia’s or Microsoft’s intentions here. Windows on ARM, yes, I can see that–but Office on ARM? What is it about Intel that isn’t good enough at running Office? Why break _so_ much compatibility, just to run Office on a different architecture. How would this change enterprise buying habits? There’s something extra up their sleeve they’re keeping secret at this stage and I expect it’s all new hardware.
Completely wild and ridiculous guess: maybe ARM has been working on adding features to the architecture to speed up the translation process from x86 to ARM? I mean, it seems like ARM has been planning this for years now, maybe they are working on something?
This is not that complicated. There are going to be more arm devices than toasters. Microsoft would be negligent to not support arm.
Personally I think it is safe to say that unless some significant breakthroughs are made in the field of software emulation, ARM has no hope of running x86 code using that approach fast enough to make users happy.
That would seem to leave hardware emulation. However, I don’t see this happening either. Its possible (anything is possible), but the problem with emulating x86 using any kind of hardware acceleration is that the part that is hard get right and make perform well is the same part that uses up vast quantities of power – the front end and decoding. If you put anything like an x86 front-end on an ARM chip it will use more power, likely a lot more power. Anyway, nvidia would have to do this part, and they don’t have an x86 license…
Some selective hardware support for emulation combined with mostly software emulation could maybe work well enough while keeping power usage under control, but I really doubt it. Transmeta did it, but it was never what anyone would call fast…
Ars had a write up that I think made a lot of sense. If you look at Office as a set of infrastructure for supporting the reading/writing of files instead looking at it as just an application it does make some sense. MS makes office into a library so to speak, its part of the OS. Its a nice and tidy way of looking at it. One possible scenario:
1. Windows is ported to ARM.
2. Office is ported to ARM (I personally expect you would see MSSQL and IIS ported as well at a minimum for server use).
3. You promote .NET as the platform API for ARM, possibly even dropping support for native code.
So what you end up with is a port of Windows where the primary API is now .NET. Anything you write for it can run on both platforms. Microsoft gets the advantage of having a single code base for both platforms, plus the added bonus of more pervasive use of .NET by ISVs (if you want to run on ARM you have to use .NET). No emulation required. This would even make a fairly good server platform (assuming IIS and MSSQL come along for the ride). Office, IIS, and MSSQL are already made to allow using .NET for extending them in a multitude of ways.
I’m not saying this is exactly what they are doing, I have no idea. But it makes sense to me. They could be going to the hard route and intend on support ARM on equal footing with x86 – native binaries, full toolset support, the whole ball of wax – but that seems more unlikely the more I think about it… It would take them multiple years to get there and I don’t think they can afford it taking that long. And ARM port with .NET as the primary API is a no brainer to me – they could accomplish that inside a year easily I think.
In a nutshell, think of Windows on ARM as a set of infrastructure for running .NET code instead of for running Windows applications.
Edited 2011-01-06 20:34 UTC
Yeah, cool, but NVIDIA already has Tegra. This is not “huge news”, it looks like just “the next generation” to me.
Tegra uses off-the-shelf cpu core from ARM.
Just when it was looking good that ARM was getting bigger and may finally bring us Linux-based netbooks and, eventually, desktops… nVidia enters the ARM processor business and Microsoft makes Windows and Office available for ARM. So much for the dream of getting an ARM-based machine in the future that does NOT come with or force the payment of the Windows tax… and so much for the added benefit of having a machine that is literally unable to run Windows.
I guess on the bright side, most of the Windows software that really matters is x86 only, and much of it will likely remain that way. Looks like even on ARM, we’ll *still* be forced to pay Microsoft only to wipe their garbage. Sad… they win yet again.
On the other hand, Windows on ARM will likely result in a surge of ARM-based computers, so not everything is bad.
Edited 2011-01-06 00:20 UTC
First, anyone over the age of twelve who uses the phrase “Windows tax” is almost certainly not going to be enlightening company.
Secondly, one man’s pain is another’s pleasure. While you clearly have deep-seated issues related to Microsoft, more than a few of us actually like their products and would disagree strongly with your opinion of them.
Personally, for instance, I can’t imagine having a machine without a copy of Visual Studio, which is – by a country mile – the best IDE I’ve ever used. Similarly, I quite appreciate having access to operating systems (XP SP3 and Windows 7) that have never actually crashed on me. That’s rather more than I can say for any of my often ill-fated encounters with Linux distros.
What should I call it then? How about, “The Additional Cost Applied to the Computer System for the License To Use The Included Microsoft Software, Which Is Hard [If Not Impossible] To Get Refunded?” Is that better?
I can’t complain as much now as I could in the past about the quality of Microsoft’s operating systems; Windows 7 is admittedly halfway decent. Now, it’s mostly their monopoly status, their forcing their OS (or at least payment) onto everyone who doesn’t choose to get even more gouged by buying an Apple, the ridiculous prices of their OSes, and their ridiculous technical and EULA limitations, etc. Not to mention their horrible customer service, outsourced to some cheap foreign country. And their “product activation” and “genuine advantage” crap that makes it almost a guarantee that you’ll have to deal with them at some point.
Oh, man. You would get bored as hell hearing about all the blue screens of death I got in XP, as well as previous versions of Windows. One standout, I recall, was one of XP’s earlier service packs; downloading torrents would never fail to cause a BSOD, and that annoying bug existed until SP2 finally came out. The only fix until SP2: Don’t download torrents or any other highly network-straining programs, or downgrade to the previous service pack. Let me just say, from around 1997 until sometime in 2006, I’ve seen Windows crash and burn spectacularly more times than I can count.
Edited 2011-01-06 01:22 UTC
Not monopoly status, ABUSE of monopoly status. That needs to be kept explicit. We don’t want microsoft stifling more innovation through strong arming vendors, running under the table deals, playing dirty with pushing “standards”, the list goes on.
I personally don’t like visual studio one bit, its pretty clunky and non intuitive and worse…forces users into its developmental paradigm. That tends to be true of most ides but is more true of VS. And then there’s the insanity of not being able to mix and match libraries at will…static/dynamic/thread mode/debug utterly maddening.
Fair enough; the abuse part is pretty important. But at this point they do pretty much have a monopoly, and they have not had to really abuse it much for years to make their OS appear on every PC; the damage was done long ago, now they just sit comfortably with their monopoly position racking up the PC sales. Netbooks are the main exception I can think of; Linux started taking off on these devices, so Microsoft raised Windows XP from its grave to trample it. Then they force the joke known as Windows 7 Starter, which doesn’t even allow changing the wallpaper, onto pretty much every netbook buyer now. Now if we want a netbook… we’re forced to pay Microsoft for a heavily cripled joke of an OS.
Edited 2011-01-06 02:17 UTC
Crippled, risen from the dead… And consumers still prefer Windows. What does that say about Linux?
If anything, it says that the makers of the netbooks made a poor choice by including some custom and awkward (even to Linux users) Linux distro instead of one of the proven, major distros. And once Microsoft realized those things were actually selling, and Vista would not be usable on them at all, XP was resurrected. Until Windows 7 was released, which was slimmed down primarily so it would run on these (but also because Vista ran like such a pig in general).
Edited 2011-01-06 03:01 UTC
Some of the early Dell netbooks running Ubuntu were broken with updates. I can dig up links if you would like.
Linux still has problems with updates.
I think I remember hearing about that, no need to dig up links.
What’s the point? Are you trying to blame that on Ubuntu (and Linux in general), instead of Dell for modifying the OS to allow certain patented codecs and other functionality, while not testing to make sure future Ubuntu system upgrades would not conflict with Dell’s own changes?
Stock Ubuntu would have likely not had a problem.
This is in the same league as those Windows XP SP3 upgrades by certain manufacturers that sent XP into an infinite reboot cycle after upgrading on machines with an AMD processor, due to a buggy third-party system file not present (or at least not as buggy) in the official Microsoft retail version.
Edited 2011-01-06 05:58 UTC
And Windows hasn’t had updates brick installs?
The real issue with Linux netbooks was that there was no standardisation at all, everyone but Dell used a custom Linux that was abandoned within months of the release of the Netbooks. Go look up how Xandros that was on the Asus EEEs was dead in the water not 6 moths out of the release, no updates to anything unless installed manually since they never kept up the repos and because it was so custom most precompiled binaries where a pita to get installed as you had to also track down the dependencies yourself.
It wasn’t a fault of Linux, it was a fault with abandonware.
you forgot about the fixing and bribery behind the whole OOXML fiasco. That happened fairly recently. Business is as usual there.
Hang on a second: presumably, you’re opposed to the idea of using Windows either because you simply don’t like the product or you find it distasteful from a purely ideological standpoint. Either of those are perfectly fine reasons not to use it. Moreover, you don’t want to buy a computer with Windows pre-installed because you regard it as a “tax” (which it isn’t) and you believe it may be difficult to obtain a refund for the unused copy of Windows.
Here’s a simple solution: don’t buy a computer with Windows pre-installed.
So, what you’re saying is… buy a Mac. Which, for a given spec, is almost always far more expensive and less-configurable than a PC–even with Windows included. And the hardware is far less configurable (in laptop territory in terms of the ability to upgrade its internals).
Add to that the fact that the whole point of spending a premium for Mac is for… eh, Mac OS X, so spending huge amounts of money only to wipe it and install another OS would make even less sense than just buying a standard PC and accepting the Microsoft payment. Then you’re stuck with some crazy vendor lock-in and forced to do everything exactly how Apple wants you to, possibly worse than even Microsoft’s OS.
Gotcha… I’ll keep that in mind the next time I go looking for a laptop or netbook.
… 😐
Simple in theory; not so simple in reality.
Edited 2011-01-06 04:41 UTC
Why does “don’t buy a computer with Windows pre-installed” sound like “buy a Mac” to you? It doesn’t to me.
If you want to buy a PC without an operating system there are plenty of places you can do so.
Honest question: Are there really any good such OEMs? Usually the ones with the “Linux” or “FreeDOS” option tend to be hard to find at Dell’s site for example, and have inferior hardware, or not the brand I’m looking for (ie. AMD instead of Intel CPUs, ATI or Intel instead of nVidia GPUs). They also tend to lack the component customization that the Windows systems have. Only by getting Windows are the higher-performance machines and extra hardware customization opened up. And I actually don’t know about any companies that sell good machines with no OS; every OEM puts Windows above all, and makes other options hard to find (even to the point of recommending *against* them when you finally do find it, just goes to show how much these companies REALLY want you to buy Windows…).
It is a nightmare trying to find a decent PC not running Windows by default with the components I want. Yet you claim they are not hard to find. Do you know something I don’t?
Hell, just look how resistant Dell was to putting Ubuntu on their machines in the first place. I’m shocked they provided it as an option on even one model, and the other companies seem just as cautious.
Edited 2011-01-06 05:03 UTC
I haven’t bought a machine from one of the major manufacturers in quite a while so I can’t comment too much on how they do things. That said, if you want to purchase, say, a Dell without Windows in the UK you can do so by getting a quote from a Dell representative on their LiveChat system. In addition, when we buy new machines at work we always have the option of purchasing without an OS pre-installed.
Personally, I tend to go to places such as overclockers.co.uk whenever I want a new system and they give customers the option of buying without any OS; there are many, many other places like this, including loads in the US.
When my mom was buying this machine from Dell years ago over the phone (which she later gave to me), I told her to ask about Ubuntu–which was somewhat “new” for Dell at the time. Ironically and not really surprisingly, they talked like they didn’t know what the hell she was talking about. The only alternative they gave aside from the crap known as Vista was a whopping 200-300 bucks to get XP instead (insanity!).
Unfortunately–the cheapest route, Vista, was chosen. And this was around 2007, 2008. I honestly have a feeling they knew, but did not want to help as it is their job to please Microsoft and sell Windows on as many systems as possible. I’m pretty sure we even asked if we could get a system with *no* OS, and they said no. We didn’t have much luck back then, and the machine has not been running Vista since it was given to me (Debian Testing right now).
I haven’t dealt with Gateway since 2001, but it seems even harder to find non-Windows machines (or even information) than Dell on their site. That is quite an accomplishment, given how much of a PITA Dell makes it. And I haven’t dealt with any other OEMs, personally.
Edited 2011-01-06 05:30 UTC
There are some OEMs that cater specifically to the OSS/Linux crowd. system76 comes to mind. There systems only come with Ubuntu installed. They are what your are looking for. I have no idea as to the quality of their machines or their support but it proves that there are OEMs that cater to Linux out there.
I heard about that company.
I’ll admit… by asking (though I didn’t specifically say it, trying to prevent blind “yeah I heard of them, they’re great just ’cause they support Linux!!! responses), I was actually hoping to get opinions on whether this company’s customer service and hardware components are good, or not. I recall their systems being at least slightly higher priced than the typical competitors’ Windows offerings (likely because of the lack of crapware and demos including anti-virus to offset the hardware cost), but that can be forgivable if their hardware and service is good.
One definite plus is that by buying from them, Microsoft doesn’t get another penny. That is great, but the company itself (support) and its hardware components need to at least be decent.
Edited 2011-01-06 07:59 UTC
In the US, check out these two companies:
http://zareason.com/shop/home.php
http://www.system76.com/
Even I know about these two companies, and I reside in Australia.
As for price versus quality … normally the higher the quality, the higher the price (altough personal computer software does seem to be a glaring exception to this general rule). In a fair market, for the exact same hardware, a Windows system typically costs (in AUD$) the following amounts more than an Ubuntu system:
(showing a DreamBook Light U11 CULV laptop model for comparison, $699 AUD for the bare hardware)
http://www.pioneercomputers.com.au/products/configure.asp?c1=3&c2=1…
Ubuntu Linux OS Pre-loaded [+$0]
Microsoft Windows XP Home Edition with Recovery CD [+$39]
Microsoft Windows 7 Home Premium (32/64 Bit) [+$129]
Microsoft Windows 7 Professional (32/64 Bit) [+$169]
All prices quoted in $AUD. $169 AUD is an additional 24% over and above the base price of the machine. Even then, the Microsoft Windows 7 Professional price is for the installation of a bare OS only, whereas Ubuntu comes with a complete set of desktop applications already included.
The exact same pricing applies to this desktop model:
http://www.pioneercomputers.com.au/products/configure.asp?c1=4&c2=2…
I am not in a position to say if the US market is a fair market, or not.
Edited 2011-01-06 10:03 UTC
That mark-up on Windows there is idiotic. These guys are probably shipping retail boxes, which is the silliest thing I’ve heard in a long time. OEM copies of Windows are a few dozens of Euros at most for small OEMs, and about 10 Euro for large OEMs.
This particular company in AU imports “bare metal” piece parts from Asia, and assembles them into personal computer systems.
http://www.pioneercomputers.com.au/products/
Nevertheless, the end products they sell are price competitive with other sources of computers in the Australian market. They have been doing this business model for 14 years.
http://www.pioneercomputers.com.au/about/
Products are marketed to Japan, China, Brazil, India, Russia, New Zealand and Asia Pacific Region.
They are a government approved/endorsed supplier:
http://www.pioneercomputers.com.au/government/
They also supply to the education market (which, in Australia, is government run).
Other companies can (and do) apply to become a reseller of their product:
http://www.pioneercomputers.com.au/resellers/
They have no less than three mass-market resellers in Australia:
http://www.pioneercomputers.com.au/buy/
Incidentally, they have some nice ARM-based tablets they call “ePads”, running Android:
http://www.pioneercomputers.com.au/products/products.asp?c1=183&c2=…
If their business model was really so silly, one wouldn’t imagine they would be this well-established and be doing it for so long.
Edited 2011-01-06 13:30 UTC
Yes, that’s right, you’re forced to pay for their OS… notwithstanding the fact every city of any decent size has dozens of small computer retailers who will happily build/sell you a computer without Windows.
If you’re too lazy to shop anywhere other than Best Buy or Staples, well then that obviously must be Microsoft’s fault… somehow.
For most of us those shops are not available, most just stare at you blankly if you start asking about anything that isn’t some useless IGP based system with Windows on it or a Mac. That or they are way more expensive then dealing with a normal OEM for the same hardware.
I know this because there are 4 of them in this town, I’ve talked to all of them, only one wont look at you in confusion if you ask about things like GPUs, TV tuners or HD capture cards, they are also by far and away the most expensive.
I don’t know, maybe you are correct, I have no company.
But it’s a Windows tax. If I bought a laptop I would more or less be forced to buy Windows. I don’t need it. It would either run OS X as hack or FreeBSD.
… or MeeGo from SuSE or something.
Edited 2011-01-06 07:24 UTC
This will certainly help ARM expand into the desktop and notebook markets.
Linux computers were available years ago. The Year of the Desktop Linux was probably some time around 2004.
One wouldn’t have happened without the other. The two stories are really the same. The nVidia desktop processors wouldn’t sell without the capability of running Windows, and Windows wouldn’t have entered the market without desktop- and laptop-capable processors entering the mainstream market.
ARM-based machines currently include a “Microsoft tax”?
Why is that a benefit, and to who? If you don’t want to run Windows, don’t. Who would want fewer choices???
Only if you buy a computer with Windows pre-installed. Fortunately, any ten-year-old can snap together their own desktop computer and install whatever OS they wish.
The theme of your topic: the only way for Linux to succeed is to have no competition. Not very confident, are you?
Linux has quite a head start on ARM, more than ten years. You don’t think that gives Linux enough advantage that they’ll be able to withstand a little competition?
Sorry I just can’t “snap together” a netbook or a notebook. It doesn’t work that way.
Nope, you can’t. A major drawback of laptops and portables, IMO.
Sure, not now, and I don’t recall saying that they already do (and if it sounded that way, it’s not what I meant). But once the ARM machines finally come into existence as actual usable general-purpose computers, good luck finding one that doesn’t come with Windows preinstalled–whether you want it or not.
I don’t run Windows. It would be a nice extension to my freedom of choice to not be forced to pay Microsoft for something I will not use, and refuse to even agree to its EULA in the first place.
No, the theme of my topic: We’re going to be forced to pay Microsoft when we buy an ARM-based computer now, just as we always have been for x86. And as things usually go–good luck getting that Windows refund. The bottom line is, I’m just getting sick and tired of paying Microsoft every f***ing time I buy a new computer! Even worse, considering I no longer run Windows–yet if I want a laptop or netbook, tough shit. ARM? Pretty soon, same thing.
Edited 2011-01-06 03:05 UTC
Its quite simple really, with MS now invading ARM space with something other then Wince they will strong arm the OEMs in the same way they have done in X86 space, this is the reason Google has gone into overdrive with Android and ChromeOS development, they’ve got to head them off before NT hits like a bomb.
While I doubt NT will be able to run well on even the 2.5Ghz quad core Cortex A15 based chips coming next year, that wont stop MS from leaning on the OEMs to carry only MS products to kill off any competition from Google or the traditional Linux distros like Ubuntu, Debian etc.
Even if NT is a turd, MS will still force it out to all the OEMs and will hit us with the same tax that you get with any mobile X86 device.
Don’t try to weasel and say DIY, ever DIY a laptop or a cellphone? Didn’t think so.
I hear you on the Windows tax, especially on mobiles, on desktops you can easily avoid it by building your own, but for a laptop you have to look for the not so well known customiser shops like:
http://www.system76.com/ and http://zareason.com/ that actually support Linux.
http://www.avadirect.com/ and http://www.pugetsystems.com/ that give Linux as a BTO option, but offer no support.
http://www.xoticpc.com/ and http://www.powernotebooks.com/ which offer you to opt out of any OS on most models, but wont install anything for you but Windows.
For ARM you have http://www.openpandora.org/ -|- http://www.littlelinuxlaptop.com/ -|- http://www.alwaysinnovating.com/ -|- http://www.norhtec.com/ who produce Linux only hardware
As well as any company pushing Android like http://www.augenus.com/ or http://www.archos.com/ etc.
…I’ll be all giddy when I go to build my next computer in a year or two… Next gen AMD, Intel or now… ARM for my CPU/MoBo? Ohhhh the humanity!
What would be REALLY interesting is OS X which probably would already run on it.
Windows on ARM? There isn’t enough NoDoze in the word for that.
This should really wake up Intel though.
Speaking of Apple (I was, I didn’t read the other posts because they are probably about Windows), they probably already know about this. I wonder if they know enough about that they will base their low end computers on it in the beginning and then later, maybe all of their computers on it.
Apple already base their low-end computers on ARM chips
Consider the iPad…
If you’re advocating an Apple desktop/laptop computer based on ARM chips, then that’s another story. Earlier, Apple managed to do the PowerPC -> x86 transition quite smoothly using PowerPC emulation, but now we’re talking about switching to a *slower* processor, so emulation is not a practical solution.
Wouldn’t all games need to be 100% .NET (or java, python, etc.) to be compatible between x86 Windows and ARM Windows?
Aren’t most games still written in C++?
One merely needs to take the C++ source code and compile it for the ARM architecture which one is targetting. One thereby produces an ARM executable binary, not an x86 executable binary, for the exact same game application.
x86/x86_64 does not have a monopoly on C++ code.
For example, most of Nokia’s Qt code is written in C++. Qt is the basis of Meego Linux.
http://liliputing.com/2010/05/despite-intel-backing-meego-linux-to-…
http://www.engadget.com/2010/06/21/nokia-7-or-9-inch-meego-tablet-r…
Edited 2011-01-07 02:05 UTC
In theory, this approach should work fine. But in practice it’s not going to be anywhere near as simple as that.
For instance, game producers could very well find themselves facing a situation in which they need to link their application against some third party library. If they’ve been linking to that library as an x86 binary, there’s no guarantee that they’ll be able to obtain corresponding binaries for ARM. Depending on the level to which they rely on that library, this could end up being a complete show stopper for many projects.
If true, this has nothing at all to do with either C++ or games, it is a completely separate issue.
Insofar as this issue itself goes … I thought that the general rule on Windows was not to dynamically link libraries. That is to say, it wasn’t advisable on Windows to simply expect another library of the correct version to already be installed on a system, so that one can link to it, but rather one included any required libraries within one’s own application code (that is, typically all required libraries are statically linked for every application). In turn, this is the reason why the same number and complexity of Windows applications requires many times more disk space than Linux applications.
It’ll be interesting to see what is actually available for this platform when it is released. Does this mean wine is going to need a rewrite to become compatible with Windows ARM applications? On day one of release Linux will have more applications and how long will it take for Microsoft partners to catch up? In terms of marketshare I’m skeptical of what this will mean for both linux and windows.
Wine Is Not an Emulator.
http://wiki.winehq.org/Debunking_Wine_Myths#head-3049e5862dfd3e6de4…
Wine right nows runs on x86 only because Windows executables run on x86 only.
AFAIK, if one were to re-compile Wine source code to target ARM, then the resulting Wine for ARM program would run Windows for ARM executable binaries on Linux ARM machines.
The Linux community has source code for the majority of Linux applications, there are relatively few closed source drivers and applications for Linux. It would be utterly stupid for a hardware OEM to make an ARM system right now for which there was no open source Linux driver for some of the components. All ARM systems will have all open source drivers for all components. Almost the entire Linux ecosystem software ensemble is available right now for ARM.
http://www.linuxine.com/story/ubuntu-based-arm-server-runs-80-watts
http://www.linuxfordevices.com/c/a/News/ZT-Systems-R1801e-/?kc=rss
http://www.computersandcomputersoftware.com/?p=1591
Windows is almost the exact reverse. Microsoft holds the source code for only the bare OS and a very few applications and drivers, all other source code for Windows is held by third parties.
This means that when Windows for ARM is first released, there will only be the bare OS itself and MS Office and a few other Microsoft applications available for it. Microsoft will have to plead with hardware OEMs to create systems with only components for which Microsoft is in a position to provide drivers.
As a consequence, when it first comes out, Windows for ARM will be very much a Johnny-come-lately that works fully only a very few selected ARM machines and with very few available applications.
Edited 2011-01-07 13:53 UTC
Not so.
This is to rather completely miss the point of what I was saying. The issue is not what libraries are available to be installed with the application when it’s deployed; that’s a secondary concern. The real issue arises when there’s a chance that needed libraries won’t be available even to the developer because the producer of a library hasn’t made them available for ARM.
I understand that everyone likes the idea of portable code as an ideal. Unfortunately, things don’t work like that in practice when you’re dealing with C++. Even something like Boost, a wonderful and almost entirely header-only library, has significant components that need to be built explicitly before use.