Vanderpool : the first HyperOS chip
According to Intel the future microprocessor will be multiple OS.
An article (sorry in french) relate an article edited in the revue New Scientist by Mike Ferron-Jones, chief of the marketing of the advanced technologies of Intel center at Hillsboro, Oregon.
Until now, the multiple system OS on a unique computer presented severe difficulties due to the memory gestion, and various drivers required for printers, and external devices.
Solutions already existed, based upon virtual software (like virtual PC for expample) who was managed memory affectations, but at the prize of poor performances.
Vanderpool refer to a new line of chip able to execute natively the principle OS.
Naturally , this project concerning in particular, Windows XP, Linux and Mac OS X, will require also a virtual software, but will improve and simplifie it's use.
The chips should execute this basic software via an Hyper OS, a miniature OS simple and stable. It will allow PC chips to acess to alternative OS without leaving Windows and others Microsoft applications.
This announce may bring some doubts in the privelegied relationship between Intel and Microsoft.
The Vanderpool chips should ship in 2007 or 2008, and should make a huge change in the micro-computer world, and should foster developpement of alternatives OS.
The big question that remain is : who will develop HyperOS ? Intel who has the necessary expertise, but who have always refused to do so ...
the question is open.
An article (sorry in french) relate an article edited in the revue New Scientist by Mike Ferron-Jones, chief of the marketing of the advanced technologies of Intel center at Hillsboro, Oregon.
Until now, the multiple system OS on a unique computer presented severe difficulties due to the memory gestion, and various drivers required for printers, and external devices.
Solutions already existed, based upon virtual software (like virtual PC for expample) who was managed memory affectations, but at the prize of poor performances.
Vanderpool refer to a new line of chip able to execute natively the principle OS.
Naturally , this project concerning in particular, Windows XP, Linux and Mac OS X, will require also a virtual software, but will improve and simplifie it's use.
The chips should execute this basic software via an Hyper OS, a miniature OS simple and stable. It will allow PC chips to acess to alternative OS without leaving Windows and others Microsoft applications.
This announce may bring some doubts in the privelegied relationship between Intel and Microsoft.
The Vanderpool chips should ship in 2007 or 2008, and should make a huge change in the micro-computer world, and should foster developpement of alternatives OS.
The big question that remain is : who will develop HyperOS ? Intel who has the necessary expertise, but who have always refused to do so ...
the question is open.
Comments
Originally posted by Powerdoc
According to Intel the future microprocessor will be multiple OS.
(snip)
Naturally , this project concerning in particular, Windows XP, Linux and Mac OS X, will require also a virtual software, but will improve and simplifie it's use.
"pourra indifféremment exécuter Windows XP, Linux et Mac OSx"
does this really mean that operating systems compiled for a non-x86 architecture will run natively at reasonable speed? or do they assume that Mac OS would be on the x86 ISA by then?
if it's true - you can see why it's going to take them another 4-5 years.
the article in english: http://www.newscientist.com/news/news.jsp?id=ns99994215
the slashdot thread: http://slashdot.org/article.pl?sid=0...id=185&tid=190
Originally posted by staphbaby
"pourra indifféremment exécuter Windows XP, Linux et Mac OSx"
does this really mean that operating systems compiled for a non-x86 architecture will run natively at reasonable speed? or do they assume that Mac OS would be on the x86 ISA by then?
if it's true - you can see why it's going to take them another 4-5 years.
the article in english: http://www.newscientist.com/news/news.jsp?id=ns99994215
the slashdot thread: http://slashdot.org/article.pl?sid=0...id=185&tid=190
It means that non X-86 or X-86 architecture will run at a reasonable speed. This chip is not build upon X-86 architecture, but is a chip who translate very fast differents codes. An evolution of the Transmeta crusoe chip if you prefer.
The goal of this chip is to make the transition between CISC and the X86 world and a Risc architecture for the consumer market.
The itanium is pretty lame for translating via virtual software X86 code. The project Vanderpool is the way Intel want to do the transition.
An hard transiton that Apple has suceed (but with difficulties) with his PPC chips.
Originally posted by Rhumgod
Pure Intel vapor ware. Some marketing yuck-yuck is smoking too much hippee lettuce
I am not a fan of Intel, a company that have good engineers, but who has for first priority marketing considerations and not performance.
However i did not remember that Intel announced many vapor ware products in the past. We should consider this new seriously, even if this type of chip is not for tomorrow.
Originally posted by Aquatic
Did God tell you that?
He doesn't really talk to me. But my dog does.
Seriously, though, are you questioning my ability to predict the future?
Clearly we won't see Dells running Mac OS X without a major shift in Apple policy (one that would probably kill its hardware sales, at least in the business and education sectors, where looks aren't important), so is Intel proposing designing a whole new range of chips just so that a 3% market share company like Apple will be able to make Macs that run Windows?
I don't get it.
My otherwise bright operating systems professor actually bought into that hype, too, which was particularly funny.
Originally posted by Powerdoc
...The big question that remain is : who will develop HyperOS ? Intel who has the necessary expertise, but who have always refused to do so ...
the question is open.
Why would you want a Hyper OS when you could recompile for the chip? It would seam to me that you would loose most of the chip specific optimizations with this Transmeta like chip, or the Hyper OS would balloon as libraries were written to take care of the chip specific code. Also, if you moved to a new chip you might take a performance hit as you lost the old optimizations, and it may take longer to develop, but in the long run your OS would run better on that chip if it's talking the "native" language.
Wired just needed a source for its next "technology will save the world" article.
Much of this discussion is just silly. There is no need for a new archicture from Intel, nor is there a need for Apple to abandon it's vertical integration model to port OS X to Intel.
Apple can produce a Mac OS X that will run on it's own machines only , with the Intel chip inside. All it takes is a recompile. Applications wil also have to recompile.
In fact all applications and frameworks could be "fat", which is to say that the binary can be in <package>/Contents/Intel and <package>/Contents/MacOS ( really meaning PPC)
As for the sheer nonsense about Apple not controlling the "whole widget", well, words fail me. Apple does not in fact manufacture it's own chips - there are 2 companies at present manufacturing CPUS for the Mac, and the other industry standard components - like video cards - come from across the board.
Here is all they have to do.
Transition the line to intel over a few years, keeping applications fat.
Port the AppKit and Carbon to the Windows side ( i.e. OpenStep, and the QuickTime API that QuickTime player and iTunes already use - mostly done already)
Install such packages as default on the Apple Windows box.
Port the iLife suite to Windows ( gnerally a recompile if the AppKit is ported) , and get it for free- or exclusively - on an Apple Machine running Windows.
Market that Apple does Windows better than WIndows. Re-start the swichers campaign.
Use the software group to make sure that the Apple Windows box has correct drivers for all Apple external products like the iPod.
Allow dual booting into Mac OS X.
This is still vertical integration.
Have a modification on the Mac OS X side for switchers ( i.e. put the close box on the right - there is no advantage either way)
This would all allow people who like Apple's hardware, and integration, and who are dithering on Mac Os X because of the cost of new software acqusitions to transition over smoothly.
Who would deny that Apple could get 10% of the market, even 20%? That is where it's mind share is at. It is the second most recognised brand in the world with 3% market share. Preposterous.
And if the Mac OS X side fails ( but why would it if it is better?), then Apple would still survive. Anyway the 3% that constitutes Apple's market now is not going to go elsewhere, they can just not boot into windows if they so wish. Or eliminate it entirely from their machine, but even they would admit that it unleashes more software opportunities, especailly for games)
That is what Apple should so, say I as a shareholder.
- How do you recompile Mac OS Classic, or hand-written ppc machine code for Intel?
- Why would Apple's users and developers put up with another major switch, this time even worse than the one from classic to OSX? Not only will all programs need to be re-released (and since when has there ever been such a thing as a 'simple re-compile' for any non-trivial application?), but for users buying new (intel) hardware there will be no means whatsoever to run their old programs.
- Why would switching to intel benefit Apple? So that their hardware users can choose to run Windows instead of Mac OS X? Is it likely that Apple wants this? If they did they would just make PC clones that looked like Macs, or Macs with dual processor boards (X86 and PPC), it would be a lot cheaper than cross-platform redevelopment of all their software.
Now if apple could find a cost effective way to make Macs with Intel and IBM chips in them, and have a smart way of integrating Windows and Mac software, a bit like they do with classic now, well that would start to look like something worth buying - hardware VPC, only better. It'll never happen though.
Socrates
- How do you recompile Mac OS Classic, or hand-written ppc machine code for Intel?
You don't. Classic is on the way out. nobody in Apple is writing ppc machine code for OS X proper
- Why would Apple's users and developers put up with another major switch, this time even worse than the one from classic to OSX? Not only will all programs need to be re-released (and since when has there ever been such a thing as a 'simple re-compile' for any non-trivial application?), but for users buying new (intel) hardware there will be no means whatsoever to run their old programs.
It's just a recompile. Since when would it be a simple re-compile? Since now. If the code is written against the carbon API, or against the APPKIt it would just work after a re-compile. All big endian issues etc. will be taken care off because Carbon and Cocoa makes the structures, or objects, opaque so that you need to use a API to get at the internals.
As for software, if you own version 6.1 of application <foo>, then the re-compile of Application <foo> , producing version 6.1.1, will be free to you on your intel box, like any update and will be placed in the App package I talked about. Apple can announce this at WWDC, and in kitchens get 90% of all applications re-compiled that week. That is a much easier job than the original porting to Carbon, of Mac Os 9 apps, which facilitates it.
- Why would switching to intel benefit Apple? So that their hardware users can choose to run Windows instead of Mac OS X? Is it likely that Apple wants this? If they did they would just make PC clones that looked like Macs, or Macs with dual processor boards (X86 and PPC), it would be a lot cheaper than cross-platform redevelopment of all their software.
I answered the benefits already. The "cost" to Apple of "re-developing" their software is setting a flag in GCC that builds for intel. That cost is exactly zero. And they probably do it already.
Originally posted by asdasd
You don't. Classic is on the way out. nobody in Apple is writing ppc machine code for OS X proper
It's just a recompile. Since when would it be a simple re-compile? Since now. If the code is written against the carbon API, or against the APPKIt it would just work after a re-compile. All big endian issues etc. will be taken care off because Carbon and Cocoa makes the structures, or objects, opaque so that you need to use a API to get at the internals.
As for software, if you own version 6.1 of application <foo>, then the re-compile of Application <foo> , producing version 6.1.1, will be free to you on your intel box, like any update and will be placed in the App package I talked about. Apple can announce this at WWDC, and in kitchens get 90% of all applications re-compiled that week. That is a much easier job than the original porting to Carbon, of Mac Os 9 apps, which facilitates it.
I answered the benefits already. The "cost" to Apple of "re-developing" their software is setting a flag in GCC that builds for intel. That cost is exactly zero. And they probably do it already.
Naw ... I don't see it, the G5 has arrived, the panic is <mostly> over:
Though out of pure speculation - I can see Apple keeping an internal version of something very close to "Yellow Box For Windows" to itself, which it likely recently updated to make iTunes for Windows, giving Apple the advantage that it can pull of it's hat any time it has to make key Windows versions of it's apps.
This way Apple controls when PC's get a killer Mac app, and which killer Mac Apps the PC gets (read: only those that make PC users want to buy an Apple product).
ta
Originally posted by asdasd
Please.
Please is right...
Here is all they have to do.
Really? This is all Apple would have to do? You should go call up Jobs and let him know you have the answer to save Apple.
Port the iLife suite to Windows ( gnerally a recompile if the AppKit is ported) , and get it for free- or exclusively - on an Apple Machine running Windows.
Um, not quite. There is this little thing called optimization, and more specifically Altivec. The thing you seem not to get is that the optimization isn't just a drop in recompile. It would take some work to "port" (or worded correctly, rewrite) any optimazations done for the G3/G4/G5 and optimize for the P4.
Market that Apple does Windows better than WIndows. Re-start the swichers campaign.
Um right. I could buy a cheaper Dell (running the same hardware as Apple) download the Apple apps (all the free ones most people talk about), and run them under Windows on my Dell. Why would I buy an Apple branded box? Why would I buy OS X, if I can get all the iApps for free on Windows? How does Apple make money again?
Who would deny that Apple could get 10% of the market, even 20%? That is where it's mind share is at. It is the second most recognised brand in the world with 3% market share. Preposterous.
Ya right, they would gain nothing, and potentially lose it all. There hardware marketshare would be less than 1% (if that high), and likewise for the OS. They wouldn't be able to survive to sell what software they do sell.
Plus, I would argue they have a larger than 3% marketshare right now in the consumer space (I looked this data up once, and posted here). The problem is that any numbers that come out are based on consumer, business, and server sales, most of which Apple doesn't play in, nor would make any difference for most developers? Should a game developer care about business or server machine sales? No, but they still look at that marketshare, because that is what Garner feeds them.
And if the Mac OS X side fails ( but why would it if it is better?), then Apple would still survive. Anyway the 3% that constitutes Apple's market now is not going to go elsewhere, they can just not boot into windows if they so wish. Or eliminate it entirely from their machine, but even they would admit that it unleashes more software opportunities, especailly for games)
So what does Apple survive on? People wouldn't be buying the hardware, you admit that OS X fails, so what are they making money on again?
Wow, your plan is so fullproof
They can add the altivec funtctionality to a new chip , if necessary. What percentage of Apples products use that specific API anyway? Altivec is massively over-rated - the windowserver uses OpenGL to transform it's windows and that will just port to the exact same chipset on the new APple machine running with an intel cpu.
iTunes has just been ported, and that does nto seem to suffer from Altivec related issues.
The vast majority of apps have nothing to do with Altivec.
For the rest, it seems that it is you, not I, who consider OS X to be a failure.
For if they port to Intel everybody (you opine) is going to run to Windows and the 3% is going to drop to 1% ( As an side : I do agree that the market share for consumers is probably higher but lets keep these figures for the argument)
Why would this happen? Would you run off to Dell because Apple moved to Intel. Why is Apple going to lose 66% of it's already installed base?
Now to get back to the "money" quote
Um right. I could buy a cheaper Dell (running the same hardware as Apple) download the Apple apps (all the free ones most people talk about), and run them under Windows on my Dell. Why would I buy an Apple branded box? Why would I buy OS X, if I can get all the iApps for free on Windows? How does Apple make money again?
I love the nerdy "um". Should be banned from all adult discourse.
To answer the rhetorical question first, Apple makes money from hardware. Which was, in fact, my point.
To answer the rest - the iLife software ( with the exception of iTunes) will not be free ( or , as I said not available, or will not run) on a non-Apple branded Windows machine.
The cheaper Dell will not have 2 operating systems, will not come with Apple's iLife suites, and will not have the industry's best components.
Despite what Analysists say, computers are not just a commodity, they are ( particularly for portables) a status symbol. Your question is like asking why would people ever buy an IBM, or an expensive house in a good neighbourhood.
Status. Which is the point of having a brand and using it properly.
Apple has the best design,innovates on hardware, is an early adopter ( usb, firewire), has a very well known brand, and looks better than the competition.
So they charge more. And get 10% of the market and my shares go up.
One last quote.
You should go call up Jobs and let him know you have the answer to save Apple.
Maybe I should. I am his boss, remember. Along with the rest of the shareholders.
Originally posted by asdasd
They can add the altivec funtctionality to a new chip , if necessary. What percentage of Apples products use that specific API anyway? Altivec is massively over-rated - the windowserver uses OpenGL to transform it's windows and that will just port to the exact same chipset on the new APple machine running with an intel cpu.
iTunes has just been ported, and that does nto seem to suffer from Altivec related issues.
The vast majority of apps have nothing to do with Altivec.
Altivec is a great stuff. Read the article of Hannibal about the PPC 970, and you will see that it's the best SIMD unit of the market.
Without altivec Apple will be dead right now. Altivec allowed to perform Photoshop, Mp2 encoding, and final cut pro with good results.
The aqua interface use both the GPU and altivec in order to accelerate the graphic environnement. Some of the features of OS X are unique and require a vast amount of power.
Right now, all the Apple computer product line is Altivec enhanced, and thus more and more optimisation will occur.
I tunes have been rewritten entirely for windows, it's X86 code and not a recompiling. Thus i can bet that some part of the soft like encoding use SIMD optimisations, like MMX or SSE.
the iLife software ( with the exception of iTunes) will not be free ( or , as I said not available, or will not run) on a non-Apple branded Windows machine.
The cheaper Dell will not have 2 operating systems, will not come with Apple's iLife suites, and will not have the industry's best components
OK the i life suite will have a 150 $ value.
Do you really think that people will buy Apple hardware just for saving 150 $ with the I apps and for "better componements" ?
150 $ do not make a huge difference of prize, not enough to fill the gap between a Mac and a PC.
For the quality of componements, most people of the PC world are more confident in the quality of a HP, an IBM rather than in a mac.
Apple will gain a higher percentage in the OS market, but will lose some in the hardware market. The problem is that the hardware market is much more important for Apple than the software.
The goal of Apple's software, is to promote the sell of his hardware, and not the contrary.
Maybe I should. I am his boss, remember. Along with the rest of the shareholders.
Well you should be also the boss of GW Bush, after all you vote ...
Apple would still be vertically integrated. Mac OS X will only run on Apple machines, so they can't gain OS share to lose hardware sales.
i am NOT in favour of licencing the OS to OEM's.
I take the point on Altivec, and if it is so important it can be added as a seperate chip. I have heard rumours about the possiblity of that.
Lets say that Apple share of the consumer market is now 5% ( to get off the 3% figure that people so detest).
I myself don't care what chip is inside a Mac, neither should you if the transistion is handled seamlessly.So the share of Macintosh loyalists remains static.
Present day window users will be added to the mix. The barrier to the switch is not the high cost of a Mac ( people who can't afford it, can't afford it - and nor can they afford a BMW). The invisible cost is the the loss of software upgrades. the biggest barrier is the non-availability of some software.
I know of 5 people: from my wife , to my brother, to about 3 other friends and colleagues who would switch if they didn't have to upgrade all their software immediately, and could handle the transition in stages.
Or could use speciality ( business) software not now available for the Mac.
They love the look and feel of the hardware, the hardware specs, the brand name ( which is very important or else everything bought would be a "clone" and nobody would buy a Dell, etc.) , and even the OS.
The software upgrade cost needs to be added to the switchers cost of transitioning, or else they sometimes ( as one of my buddys is going to do) keep their old PC around for some software use. That too is a transition cost, as they cannot offset the cost of their old computer against the new.
So a dual booting PC/Mac would do the business. To at least double the market share. There is a huge demand for this - look at how linux i s spreading. And for people who do not detest Windows, even they like options, and a dual booting Mac gives them options.
In case Windows goes crazy with a subscription model.
This is the classic non-zero sum game.
And as for whether these people would transition to OS X fulltime, after a period of time on the dual booting machine - that is up to the OS.
It seems that some of the defenders of OS X seem to think that if you take the iLife suite away , the system offers nothing: I disagree.
Try expose. The Finder. The Dock . Name your favorite here. More will come in the next 2 releases before longhorn, and Apple is powering ahead.