If Windows Came to PPC, Would You Switch? 906
An anonymous reader asks: "This question was posted on Ask Slashdot about a week ago: 'If Mac OS X Came to x86, Would You Switch?' This makes me ask why not have Windows run on PowerPC? Windows/PPC would not necessarily have to run on Apple hardware, or at least not exclusively on it. I'm sure their friends at IBM and Motorola would be happy to provide chips to anyone that wanted to make computers to run this new OS. Microsoft could dust off the code from NT4/PPC, add some code from Virtual PC to get Windows/x86 compatibility, and have it up and running in about the same amount of time it would take Apple to get Mac OS X running on common Intel hardware." An additional question comes to mind, however: If Microsoft made this move, how would Intel react?
Cost? (Score:3, Insightful)
This is already an option (Score:2, Insightful)
Intel's reaction (Score:5, Insightful)
This is hilarious! (Score:2, Insightful)
I mean, those that use PPC (mostly Mac and PPC Linux users) use it becasue they don't want to use Windows. What conceivable reason would they have to switch to Windows? Hell, what reason would M$ have to port Windows to a platform where they know that no one will buy there product.
This is just dumb. Nothing to see here, move along.
In a word... (Score:5, Insightful)
Already done. It's called Microsoft Virtual PC. (Score:2, Insightful)
muuuh. (Score:5, Insightful)
I think most Windows users (myself included) don't care what hardware they use, as long as it's fast+cheap and all their apps/games run on it. I doubt that a PPC platform would be much faster/cheaper than x86 (even if you did magically manage to port Windows to it at full efficiency), and if it was, Intel/AMD would change so that it wasn't.
To sum up: I'd switch if there was a point. However there doesn't seem to be too many points.
The reason the OSX on x86 discussion came up is because people want the OS they think they want on the hardware they know they like. Asking a bunch of Linux nerds if they want to run the OS they don't like on the hardware they aren't entirely familiar with isn't going to provoke a huge discussion.
Why would anyone think this would happen? (Score:5, Insightful)
Not possible. (Score:2, Insightful)
I Used To Think So ... (Score:5, Insightful)
Why? (Score:1, Insightful)
Virtual PC (Score:2, Insightful)
Re:But why... (Score:2, Insightful)
Remember Windows NT for Alpha? (Score:5, Insightful)
The biggest thing Windows has going for it is the massive number of existing applications. But a different processor architecture would require porting. But unless the platform catches on, noone is going to port.
So why would anyone switch? This is pretty much the fate of the old Windows-Alpha port. Very few apps got ported (PuTTY is one of the few I know). Besides, most people were using Alphas as server machines, for which the software they needed was already available on the competing Unixes.
So.. no.. I don't think Windows could ever haul itself off the x86 platform. Too many legacy apps which are x86-specific.
Only if the PPC were commoditized (Score:3, Insightful)
analogy (Score:4, Insightful)
Switch? Not entirely... (Score:4, Insightful)
Re:Cost? (Score:5, Insightful)
Of course the question mentions that this question was asked and answered in the past, when IBM produced PowerPC machines that ran WinNT. Notice that there are no such machines (or OS) being produced anymore. Not enough people found the hardware to be an advantage to make it fly.
History would probably repeat itself (Score:3, Insightful)
You are obviously aware that they tried to make a go of NT on several other hardware platforms already. In addition to PowerPC there was also MIPS and Alpha. If I remember correctly, MS was dropped by one vendor and the other two were dropped by MS. There just wasn't enough of a demand for NT on workstations to pay for the development even with the cash cow of Windows on x86 PC's. So I guess my question to you is if they failed before what makes you think they could do well now?
Good news/bad news (Score:2, Insightful)
Windows users would enjoy big a big boost in security because most of the exploits for holes in the OS wouldn't run on the new architecture.
The bad news:
None of their apps or device drivers would run either.
(OK, maybe most of the apps would run under emulation, but that's never going to be particularly fast or trouble-free.)
Why? Nobody did the first time around. (Score:3, Insightful)
It didn't succeed then, it sure wouldn't now.
OTOH, I wouldn't mind if I could get a commodity PPC platform to run, say, Yellow Dog Linux on. The x86 architecture um, how to put this delicately, leaves something to be desired.
Strange assumptions... (Score:2, Insightful)
You're making it sound like IBM is really eager for someone to ask for their PPC processors. Well, guess what? Apple has been practically begging IBM for enough PPC processors for their computers, and the processors were in such short supply that they had to repeatedly delay critical product launches for many months. It happened with the new G5 Imac, and it happened with the bigger G5s.
Re:Not without Apple (Score:3, Insightful)
Apple originally switched to PowerPC because they thought it was going to become a popular "PC" chip outside of the Mac world. They wanted larger economies of scale than the old 68K line had, and they thought that Windows NT and OS/2 was going to bring that. They were wrong of course, and PPC became mainly an embedded chip.
Larger customer base for PPC => More investment in the architecture => Apple not falling years behind in hardware specs like with the G4.
Besides, if Microsoft and IBM decided to bring out Windows for PowerPC again, there probably is very little Apple can do to stop it.
Remember Windows NT for PPC? (Score:4, Insightful)
So, no, no one did care when Windows came to the PPC last time, so I doubt they'd give a flying fuck now, either.
Re:Intel's reaction (Score:2, Insightful)
Of course Intel is also developing software for Linux, like their C/C++ compiler.
And to really show who is their daddy, they develop graphics drivers for Linux.
Re:Not possible. (Score:3, Insightful)
In any case, the demand won't suddenly jump to equal that of the Wintel market so they'd have time to ramp up.
The fact that the PPC production isn't the equal of x86 doesn't mean that it isn't feasible. How long did it take AMD to gain a foothold. Or the clone manufacturers, back in the early days of the x86 PC?
Re:Cost? (Score:5, Insightful)
And of course, having Windows on PPC would probably sell more chips, creating lower prices (of course, this is in theory...:D)
Blake
Re:muuuh. (Score:2, Insightful)
Pick any two
Re:Why would anyone think this would happen? (Score:5, Insightful)
However, at the end of the day these platforms couldn't run the software people wanted without jumping through hoops like Digitals binary translator. No apps, no interest.
Re:No (Score:5, Insightful)
And the quality and polish of Apple's hardware is the other 10%. The processor architecture is of zero concern, except maybe as it pertains to battery life and heat.
Here's What I See (Score:5, Insightful)
The main question is, switching to Windows from what?
If I have a PPC and I have to run Linux, I might switch. I REALLY like Linux, but the fact is that Windows "just works" a little bit more, and while I do most of my gaming on consoles, if the games appeared, I would seriously look at buying a copy. For all our complaining, Windows does have a lot going for it. I could always dual boot anyways. A true copy of Office could come in handy.
If I have a PPC and it's a Mac with OS X... I don't see why ANYONE would. It's got the great design of the Mac and stability and CLI goodness of Unix. And OS X already HAS Office, so that point is moot. The only thing that I could think of would be the games, and Apple could push more on that (better hardware (GFX cards not 6-12 months behind x86) would help). Dual boot, MAYBE.
From Linux, decent chance. From OS X, nope.
That's how I see it.
Re:Cost? (Score:3, Insightful)
Of Course, this won't happend overnight. Many don't see the reason as well. From the control industry's stand point, however, many PC's basic features is here to stay:
There are only a few companies designing industrial strength softwares and many of them are building their foundation on MS's architecture. Reason? Simple, customer wants it. Why does customer want it? Simple, it is easy to maintain (or at least people with less expertise can do some part of the maintainence). It is easier to find some different venders for the WinTel combo and negotiate prices than anything else out there. Now that you have many pieces out there, why re-invent the wheels when you can get a reasonable price licensing other's software?
So why bring PPC in? PPC right now can be more expansive, but I believe it is a much better architecture. It also runs cooler, compare to the 3+GHZ Intel chips. Power consumption is key since not only do you have to pay for electricity, but you also have to pay for backing up those power usage in control environment. The few PC based control system I've seen try to use mobile cpu to achieve the balance of speed, faster development, and power usage. Howver, I believe with adequate R&D PPC can reap more benefit balancing all three aspects. x86 is not designed to conservatively use power in principle. It is designed to save some power(yeah right!) only because it has to. Why does company want a 550W power supply for each PC in the future? This simply don't make any sense to me. My Desktop uses more power than a halogen lamp and it is still allowed by the city ordanance? give me a break. I want my lamp!
Intel seems to see this and decide to walk away from higher GHz for now. But afterall I think PPC has a good chance to prove people it works well for them.
Silly idea it seems. But it CAN have significant impact. You just never know.....
Re:Obligatory Quote (Score:1, Insightful)
Re:How to put this... (Score:5, Insightful)
NT4's PPC port ran on RS/6000 workstations, not plastic cased consumer hardware. I ran it, on a lark, on an RS/6000 Box for a short period, before reinstalling AIX.
It was a cold desolate world out there. I had Windows NT and the default IE 2.0 web browser. I couldn't find a single other program that would run on the box. It isn't like NT4 and Alpha, where DEC developed an emulation layer to run x86 binaries on NT/Alpha. There wasn't a Damned thing, anywhere online for NT/PPC.
Ummm.. (Score:3, Insightful)
Re:Dual Boot? (Score:3, Insightful)
I think you're missing something. The games still won't run. They're compiled for x86 CPUs.
The problem isn't the operating system, it's the CPU.
Re:This is hilarious! (Score:2, Insightful)
We're talking like grownups here, i.e. a PPC chip built into some sort of system and a Windows port to run on it. Not platforms jocks flaming each other on *.advocacy newsgroups.
And I suspect a lot of Apple PPC users (and admins of RS/6000 boxes) would be mighty offended at the notion that their choice was essentially an anti-Windows one.
From a PPC fan... (Score:3, Insightful)
I run PPC, and let me tell you... If M$ started allowing fat binaries out of .NET for PPC and a significant number of programs started appearing for PPC, and they made a version of Windows that could be used inside OS X much like OS 9 or X11 are, I'd actually give money to the beast for the first time in a long while.
Now for why it won't happen... Companies would stop programming for the Mac. They'd only program for Windows, saying, "well, it runs on Windows for PPC, so get that!", and then the entire Apple platform would die out. Then Microsoft would be a near-total monopoly again (except for Linux being there, of course...) and then they might actually lose in an anti-trust case. Microsoft would then be broken up and slowly die against Linux. Well, slowly, but less slowly than they already are. This situation alone will prevent NT for PPC from ever coming back.
Oh so many people missing the point. (Score:5, Insightful)
You know, this question really doesn't have anything to do with Apple. It's a hypothetical question based on a processor architecture, and not necessarily Macintosh-based computers. Both IBM and Freescale sell Power PC microprocessors, and technically any motherboard manufacturer can design a board for a PowerPC, and buy the CPUs from either manufacturer, much as how they currently design boards for either Intel or AMD processors.
Why? Well, because the Power PC architecture doesn't have all of the nasty cruft that Intel-based systems have. Like IRQ nastiness that people keep designing around. Or the fact that they boot up in real mode, and need to be switched into protected mode as part of the boot process. Or all of the various BIOS limitations, like the fact you can't address beyond the first 1023 cylinders of a hard drive during IPL. Of the . Or the x86 instruction set and registers. [win.tue.nl]
The cost of this cruft is both cost and power. As cheap as Intel-based hardware is (due to the economies of scale), it could be cheaper if it didn't have to contain hardware and code to work around the many limitations of the architecture. It would also be quite a bit faster than it currently is.
Windows on Power PC would be a boon for users, if either (or both) IBM and Freescale could ramp up production sufficiently, and if every Intel Windows user were willing to give up their current software investments (or if such a Windows system run Intel binaries).
Of course, Windows itself would still suck :).
The things keeping people from making such a move aren't technical -- they're economic and social.
Myself, I'm composing this on a PowerBook G4 running Mac OS X. I have little or no desire to run Windows on any architecture. I doubt if you'd find too many existing Power PC users who wish they could run Windows as their core OS -- it's Windows users who should want to run to run their OS of choice on an affordable Power PC architecture.
Yaz.
Windows' compatibility is not present on PPC (Score:3, Insightful)
I believe that the one solid merit of Windows is its compatibility, just like the customizability of Linux and BSD, and the user interface of Mac. If it was not avaliable, I see no reason to use Windows at all over Mac OS.
It's already x86 (sort of). (Score:2, Insightful)
Duh, that's why I run FreeBSD.
Seriously, when the KDE is "industrial strength", there would be no need to move OS X. And that should be soon (please?)...
Re:cool... (Score:2, Insightful)
That, and it's hard for a manufacturer to get into building some Desktop PPC motherboards when the market is so small. People who want PPC either want it to run MacOSX, in which case they'll get a Mac, or they want to run Linux, in which case they might as well get a Mac or PC. Back in the day, there was the BeOS which ran on PPC hardware as well, but we all know how that ended
As for Amiga, well, I haven't heard from them in a while... I wish them the best! We really need a new player here
PS: In theory, OSX will run on pretty much any PPC hardware with the proper hacks... It can currently be done using MacOnLinux, but it'd be nice to get around that
Again, BINARIES? (Score:3, Insightful)
If they could make it fast enough and use the graphics card...fast enough that a 2 GHz G4 can emulate at LEAST a 1 GHz P3, or a dual 2.5 G5 can at least outrun a 3 GHz Pentium 4, and can use the graphics card, it might be worth switching, since you could play most Windows games and run most programs (even ones that use the graphics card).
Additionally, it'd have to be able to dual boot with OS X without a ton of work.
Re:Cost? (Score:5, Insightful)
First off, I think 'no contest' is an exagerration. There's a premium on the Macs, for sure, but it's really not such a huge one. By the time you add on the stuff that comes standard on the Mac but not on the competition the margin is a lot smaller than it used to be. And that's in the desktop realm - for laptops, Apple actually seems to have the advantage these days.
Secondly, you're completely right that the problem they're facing is one of volume, 'economy of scale.' Apples production is way too small to compete with the x86 world there. But, they've gone more and more to things like PCI and AT disk drives lately, which mitigates that to a large part. Many of their components do come from the commodity hardware world these days, and benefit from that economy of scale. Mostly what's left is the processor. And with IBM using the PPC chips in more products, with Linux working well on them, even the production of PPC chips is starting to come around - it's not just Apple using them, and the volume is growing and shows every sign it will continue to do so. At the same time, the x86 world is stagnating a bit - most folks in the western world that need or want a computer have one, and there's really no rational reason to upgrade - any machine made in the past 5 years is 100 times as powerful as it really needs to be to handle the average users demands anyway.
So I don't think the economy of scale problems, and hence the price problems, are nearly as big right now as they have been in the past, and I think they're getting smaller, not larger.
That said, if they ported Windows to PPC it wouldn't make me switch to windows. Would it make me switch to PPC? I already have a mixed bag, one Intel, one AMD, one PPC. If a Windows port to PPC resulted in increased volume for IBMs production lines, that would result in greater economy of scale and thus lower cost, and increase the odds that the next box I'd buy would be PPC. But I'd sure as hell never put windows on it.
Only if PPC were cheaper than x86 (Score:2, Insightful)
Of course, if someone started selling PPC CPUs for less than x86, AND PPC mainboards were not selling for more than comparable x86 mainboards, AND if the commercial software I use were available (We all know that any useful OSS gets ported to everything anyway...), then I might consider it.
In other words, NO.
Re:Ick, no! (Score:4, Insightful)
We run the latest version of OSX 10.3.5 on an upgraded (RAM and HD) old purple G3 iMac. We don't do games, graphics or video, but it works great for browsing the web, the kid's homework and as an iTunes server for our home network.
Re:Again, BINARIES? (Score:2, Insightful)
XP on PPC (Score:2, Insightful)
I've used NT 3.51 on MIPS/ARc, and I must say, it's a lot like using NT 3.51 on a 486. Whoopdedoo. I doubt there would be any difference today with XP on PPC. Other than the vital, unrelenting clueless of a large percentage of Microsoft's userbase. Can you imaging how many store returns, support calls, and other costs would be incurred every time a comsumer bought product X for the PPC and tried to load it on their PC.
And emulation. Fah! Emulation, bintrans, dynamic recompilation. Neat shit. Not a consumer grade item.
There is absolutely no economic advantage to a Microsoft PPC move. If Apple ever starts the long-rumored downward spiral, they may try to bail out with full commoditization. And may have ported the OSX gui to x86 + Darwin waiting in the wings. That is to their advantage.
Evidence that Microsoft would never consider this move. The XBox. For reasons of time and monetary constraint, Microsoft chose to hack together an x86-based gaming console instead of porting to a custom CPU. Likewise with the embedded market. I don't know if anyone thinks XP embedded is worthwhile (esp. with VXworks, Lynx, and QNX to compete with), but Microsoft didn't trouble to port a Windows subset to any other BSP's either.
We already have a test for this question. (Score:4, Insightful)
Re:How to put this... (Score:4, Insightful)
.NET apps are compiled to an intermediate language (IL) that is portable across architectures, like Java bytecode. When the program is loaded, the IL is compiled into native code on demand (by the JIT). Currently, executables also contain a pre-compiled version for x86.
I'm not saying by any means that Microsoft is first at this or that their method is perfect: only that MS does in fact have a plan to make it easy to create cross-platform apps where the developers don't have to worry about the target platform much.
Re:But why... (Score:3, Insightful)
Otherwise it's a total non-issue. Data is loaded by cache lines that are 64 or 128 bytes in size in pretty much all modern processors. If there ever was any difference in performance from big endian vs. little endian (there wasn't except for above-mentioned incompetant-monkey code) then this completely erases it.
Eek - abstract questions! (Score:2, Insightful)
Interesting question, yes. A little 'speculative' though.
non (Score:3, Insightful)
On the other hand, there is no benefit at all (none, zero) to running any version of Windows on an Apple chip. Windows would be equally as bad running on an expensive G5 as it would on any other chip.
This is like asking "Would people buy a Jaguar designed and built in Detroit even though they remain as expensive as ever?". The answer is no, not many will, and fewer will like it.
No... (Score:1, Insightful)
isn't this vapourware? (Score:3, Insightful)
Re:Obligatory Quote (Score:2, Insightful)
No, the market decided that it did not want a multiplatform Windows...
Re:How to put this... (Score:2, Insightful)
Cost, Performance, why buy a Mac? (Score:5, Insightful)
IF and ONLY IF you want precisely that set of "stuff". I don't want a built-in monitor on my desktop Mac, and I don't want a huge dual-capable tower with multiple cooling zones either. I already have a better monitor than Apple ships on the eMac or most of the G4 iMacs, and I can't afford a flat panel.
So to buy a new Mac to replace my Beige G3 (the last Mac Apple made that had the kind of tradeoffs I'm looking for) I would have to pay a huge premium for "stuff" I don't want.
for laptops, Apple actually seems to have the advantage these days
I can get a new 15" 1024x768 Windows laptop for under $1000. I can get one with a 15" screen that can display 1280x1024 for the same price as a 14" iBook with a 1024x768 screen. If you look at the 15" Powerbook I can get an IBM Thinkpad with the best keyboard on the market that'll display 1400x1050.
The advantage to Apple, and what keeps me using my upgraded Beige G3 (G4/466 + Radeon 7000) instead of the 1.7 GHz P4 Intel box I "downgraded" from (and that cost less than this pre-iMac Mac and its upgrades) is the OS. I can't imagine why anyone would want to run Windows on anything but an x86: the whole point to Windows is the applications, and even the best just-in-time recompilers aren't going to make anything but a real x86 cost-effective.
Trust me on this, I've still got an ARC-console Alpha in the lab at work. DEC's recompilation technology was insanely great, and the Alpha is a wonderful target for recompilation because of its low overhead instruction set and massive register bank, and it wasn't enough to make an Alpha run x86 code competitively.
I can't imagine buying a PPC-based machine to run a Windows desktop (the XboX 2 is a different story, of course, again because of the applications). Mac OS X makes the price premium (the very real premium) worth it, but spending more to run Windows slower? I don't think so.
If Windows came to Itanium, would you switch? (Score:3, Insightful)
Think about it. Intel has spent YEARS pushing the Itanium architecture harder than anyone could possibly afford to spend pushing PPC. HP's Itanium boxes are finally getting to the point where the idea of them taking over from Alpha and Precision Architecture on UNIX servers isn't actually insane. How much impact are they making on the desktop, after all those years of aggressive promotion?
They've been so unsuccessful even getting people to think seriously about it that AMD's been able to steal a march on intel with a 64-bit extension to the x86 architecture.
Windows and the x86 are siamese twins. The only way to get people to switch from the x86 is if you can get an OS, chip, and emulation architecture that lets you run x86 code faster than the real x86, cheaper than the real x86, and cooler than the real x86... and if you do that, you'll sell it as a new implementation of the x86, as Intel, AMD, and everyone else who's built an x86 on a RISC/VLIW core has already done.
No need here (Score:3, Insightful)
The problem with this idea is that the people who are interested in Thinking Outside the Intel Box are also, by and large, the people who are interested in thinking outside of the Microsoft box. And they already have a choice of solutions. One more is no big deal.
Microsoft would also have to overcome the ill will they generated for themselves with their last foray into PPC territory, which ended with many customers left twisting slowly in the wind. How long will we be supported *this* time, will be the question of the hour.
Re:Obligatory Quote (Score:3, Insightful)
nobody will switch for that mich of a price increase.
get me $98.00 motherboards and $100.00 processors and I'll give it a go.
but right now PPC and Alpha motherboards are insanely priced and you have to get specalty shops to sell you the processors at a premium.
My point of view: (Score:2, Insightful)
Unfortunately, the result is the PowerPC platform today has absolutely no inherent advantages. Nearly every modern processor available for the desktop has integrated optimized pipelines and superscalar execution units, plus a whole slew of other advanced features designed to streamline execution flow.
PowerPC, in that same timeframe, has also borrowed many features from the market, and is now much changed from when the first cores were released. But this is not an attempt to revolutionize, so much as it is an attempt to keep pace and stay competitive.
The end result: you have a whole range of chip selections to choose from today, and they're quite varied in what they offer. The PowerPC 970FX stays strictly in the midrange: it has competitive but not leading performance, good multi-CPU scalability, and has a combination of slightly lower power usage and on-the-fly frequency and voltage adjustments.
The thing is, you can find the same features in x86 processors, they're just spread out differently. You can typically get more performance for your dollar from chips from Intel or AMD. AMD has also seen similar improvements in power consumption in the move to 90nm as IBM has seen for the 970FX, so there's more than one competitor offer a good balance of power and performance. The Opteron also offers the most efficient multi-processor bus subsystem ever to grace a low-end server, so there's definitely still competition there. Finally, both Intel and AMD offer chips with on-the-fly voltage and frequency scaling, although it's still not an easy task to find a desktop board for the Pentium M (that will change soon).
So basically, with no performance advantage, Windows on PowerPC would flounder...and hey, it did flounder, way back in the NT 4 days. Funny that, the RISCy Pentium Pro was the final nail in the NT 4 PowerPC's coffin.
Not preloaded? (Score:2, Insightful)
Right now, the only PPC machines shipping in serious quantities are from Apple, and they already have a better OS, so why would they preload Windows? It just doesn't make sense.
So what PPC machines are there that would/could come with Windows preloaded? No big producers that I can think of, just fringe stuff like AmigaOne, etc.
Re:Obligatory Quote (Score:4, Insightful)
Sad that it'll be the fastest I'll ever see Windows go, possibly for a few decades. Not even the quad Xeon CPUs nor the Itanium IIs we have here are as fast as the Alpha was (in perception, not clock speed). I mean, I'd login and *BAM* the desktop was right there, ready to use. Dialogs and windows were displayed instantly in response to my input.
This Alpha dated from 1997. The best x86 boxes I've used still are dog-slow perception-wise as a result.
Re:Cost? (Score:3, Insightful)
I think the economy of scale is what works in Apple's favor with laptops. They're a leading manufacturer of laptops, so they have a volume at least on par with any x86 laptop company. When it comes to portables, commoditization is less of a factor. Laptops are mostly "custom" hardware that's specific to a model. As such, especially considering their limited variety of products, they get as much of a benefit of volume as other companies do. In the desktop market, however, there are very distinct advantages to x86 parts.
In addition, the more-efficient nature of the PPC platform has to work in Apple's favor when it comes to things that force other companies to engineer more expensive solutions, such as dissipating excess heat and providing extra battery power.