Slashdot Mirror
Will Vanderpool Make Linux More Popular?
Digitaldonkey writes "New Scientist is reporting that Intel's forthcoming multi-core processor architecture, codenamed "Vanderpool", could undermine Microsoft's dominance by letting other operating systems run simultaneously more easily. From the article: 'The chip will allow future machines to run, say, Windows XP together with Linux or the Apple operating system as easily as today's Windows computers run Word and Internet Explorer simultaneously.'"
Sounds similar to what IBM does with the AS/400 - allowing hardware subsystems to run different operating systems.
A fine is a tax you pay for doing wrong and a tax is a fine you pay for doing all right.
There are already ways to run Linux on a windows machine, and visa versa.. (VMware comes to mind)
And with todays already beefy processors, it runs pretty good, albeit not perfectly..
It seems this would only impact the share of people who are already using VMware to do this sort of thing..
Who knows
Excuse me, I don't mean to impose, but I am the ocean
The article tries to make it sound so new. Mainframes and high end systems have been able to support multiple OS for a while now.
If you'd take a look at the graphic from the article you will see that there is some sort of "OS" indeed below Linux/Windows/whatever. They call it Hyper-OS.
If you read the article you would know that this does not eliminate the need for virtualization software (Virtual PC) it just makes it run more like an application as opposed to a resource hungry OS. Similar to what Mac OSX does with classic mode.
Certain properties of the x86 architecture make it a hard chip to "virtualize" (sort of partition the processor into two virtual processors), which is what VMWare does. Chips can be designed specifically to be easily virtualizable, making applications like VMWare almost trivial to code while being much, much faster. If Intel does somehow retrofit virtualization capabilities onto a x86 chip, it could be a big boon for Linux. An open-source VMWare clone could be written quite easily, and it would run Windows almost as fast as it would run natively.
main(c,r){for(r=32;r;) printf(++c>31?c=!r--,"\n":c<r?" ":~c&r?" `":" #");}
Ever heard of Rhapsody DR-2? It's the last public release of Mac OS X on the PC. Apple actually keeps a Darwin x86 port somewhat up to date. They just don't keep Aqua up to date...
Stop posting this irrelevant link.
Darwin is a Mach based unix, on top of which one can run X Windows. It is *not* Mac OS X. Specifically, The Aqua user interface (which all native Mac OS X apps use), the Carbon APIs (which legacy Mac apps, like Internet Explorer, and Photoshop use), and Cocoa, (which newer Apps such as Mail and Safari use), are *not* open source.
Aqua, Carbon, and Cocoa are *not* part of Darwin. So, no, you cannot run Mac OS X just because there is an x86 version of Darwin. You can run yet another *nix on x86 with Darwin, but you cannot run Mac OS X.
Are people really this misinformed? How did parent get modded up?
There's also inaccuracy like this:
"For example, when an OS queries a Pentium processor about the total memory available, the chip normally informs the OS about all the free memory [...]"
I don't think a processor chip in itself is sentient enought to inform the OS of anything... It would actually be more understandable (less confusing) to the layman if they made clear the division of work between OS services, OS kernel, and BIOS.
As an aside, I wonder why nobody talks about BIOS development when new CPU architectures are discussed. Are we at perfection already or what? What's the scene with open source BIOSes? Surely BIOS is a factor in efficiency and raw performance? (IANAEngineer.)
Disclaimer: I work for Intel.
Vanderpool is the codename for Intel's hardware virtualization technology. It is independent of and different from Intel's plans to put multiple processor cores on a single die. You do not need Vanderpool technology to have multicore, or vice versa.
I do not speak for Intel. My opinions are not necessarily those of Intel's.
Jonathan Pearce jonathan@pearce.name
3EAAFB2A http://www.jonathan.pearce.name/
"Just like Macintosh did in 1984. And just look how many people swarmed to its easier-to-use machine. "
Umm, pretty much everybody. You appear to have not been around in that era, but the vast majority of users at that time used a command line shell (DOS or *nix), and openly laughed at the whole idea of a GUI.
What are they running now? That's right, a GUI, copied pretty much exactly from the Mac OS.
And, no, the Xerox Alto did *not* operate like the Mac OS - have you ever seen one in use? Apple paid Xerox for the right to use certain features of the Alto, but the Mac OS was a complete, ground up redesign of the WIMP GUI. In fact, so much so, that people nowadays think that all GUIs have always been like this.
However, GUIs were not like they are now until the Apple design team, led by Jeff Raskin, created the Lisa and the Mac. If you use Windows or X Windows, you're using the WIMP GUI first designed by Apple.
Disclaimer: I work for Intel (in an area having nothing whatsoever to do with Vanderpool), but the comments here are my own personal opinion. That said...
y notespeakers.htm#tuesday
I saw a demo of Vanderpool at Intel Developer Forum last month. In the demo, the system with a single processor was simultaneously running some version of Windows playing a media clip (a Simpsons episode) while at the same time on another monitor, another copy of Windows was running and was rebooted in order to update a device driver. The video clip played on.
My take on this (having never heard of it before I saw the IDF demo) was some sort of hardware-assisted VM. It is definitely nothing to do with multicore, as another Intel compatriot noted here.
You can read the transcript of Paul Otellini's Keynote where he presented Vanderpool at http://www.intel.com/idf/us/fall2003/conf_info/ke
I don't know if there were specific presentations on Vanderpool Technology at IDF - if there were, you'll be able to find them at http://www.intel.com/idf/us/fall2003/index.htm after November 2.
Vanderpool technology is not tied to a multi-core implementation. It's a set of extensions to x86 that make virtualization easier.
Currently, programs like VMWare need to play some extremely ugly tricks to get virtualization to work due to various issues with x86. This technology will make life easier for those wanting to virtualize the CPU.
So, just to be clear... Vanderpool and multicore are completely orthogonal.
While it's hard to diagnose your problem without more specifics, perhaps a trip to the knowledge base would help. You mentioned that numerous systems simultaneously sustained the same problem. Perhaps your 300-person company should look into what was deployed when everything broke. It doesn't take a logician to tell you that's the best place to start when diagnosing your problem.
This allows the CPU to schedule and dispatch a virtual system (in its chosen architectual mode and configuration) with a single instruction. Execution under SIE continues until the end of the dispatch timeslice, or intervention is required from the hypervisor. This dramatically simplifies the operation of LPAR or z/VM. The instruction takes a 4k descriptor block as a parameter, which describes all the attributes and configuration of the virtual system.
Essentially, when IBM created the SIE instruction, they moved the bulk of the (pre-SIE) VM operating system into microcode. This is an extraordinarily powerful facility. We can only hope that Intel has something similar in mind.
More info on SIE here: http://www.research.ibm.com/journal/sj/301/ibmsj30 01E.pdf
Can You Say Linux? I Knew That You Could.