With hyperthreading, the OS recognizes one physical processor as two CPUs, with nothing more than standard SMP behavior.
In fact, the Linux scheduler had to be tweaked in order to weaken restrictions on passing processes from one logical CPU to another, if both were on the same chip. Otherwise, hyperthreading actually caused a performance loss when treated as an SMP system.
The same arguments were abound when Intel introduced HT. And Microsoft didn't touch the licensing.
My guess is XP Home will continue to only use one logical processor, while XP Pro will use two. (Now, whether the "second" logical processor is HT on the first core, or primary use on the second core, remains to be seen.)
That's funny, 'cause before the McDonalds down the street from where I worked turned off the order-confirmation screen a few months ago, it'd been showing a Windows GPF error for several weeks...
Requires stability. Indeed.
I told them to power the machine off and back on, and they said they'd tried that. They were still waiting for someone from McD's IT department to come down and fix the box.
The problem I see is the difficulty in producing a machine that can not only reproduce itself, but do something useful as well.
I think a much more likely scenario would be producing a solution consisting of two machine species. One species does useful work. The other species makes copies of the first species.
The two species would depend on different keyed source chemicals. If you wanted to increase the rate at which you build load-bearing bots, you increase the availability of one chemical. If you wanted to increase the rate at which you were doing useful work, you increase the availability of the other chemical.
If you run Linux, I'd wager that the source code was available for almost all of the software on your machine. Meaning it can be compiled with optimizations for your CPU.
With Via's CPUs, some functionality, like encryption and random number generation, are supported by the kernel, which then makes the functionality available to userland in ways that don't require the software to explicitly support Via's CPU.
I was referring primarily to AMD versus Intel CPUs. Not only do they support different instruction set extensions, you need to provide your instructions at different intervals and orders in order to get the most out of the CPU.
Not to mention x86-64...
DVD install media only holds 4.7GB, there are several distros which are already approaching (or have already passed) this if you add up all the CDs...
All of "testing" Debian requires 13 CDs, or 3 DVDs. I've got the 13-CD set.
Microsoft has the advantage that most of their end-users got Windows on their machines pre-installed. That means they can include a lot more data in their OS, which means they can add code optimized for the popular processors.
I wouldn't be surprised to see similar efforts in the Linux community as, for example, DVD install media becomes more common, but there's still one problem. gcc doesn't make optimization for multiple systems easy. mplayer does it, but that's all I know of.
Slashdot Mirror
Picture a new kind of beanie...instead of a propellor, it's an antenna for your cell phone.
I don't think active support will be required.
With hyperthreading, the OS recognizes one physical processor as two CPUs, with nothing more than standard SMP behavior.
In fact, the Linux scheduler had to be tweaked in order to weaken restrictions on passing processes from one logical CPU to another, if both were on the same chip. Otherwise, hyperthreading actually caused a performance loss when treated as an SMP system.
I'm not sure how the system is normally supposed to work, but you might want to be careful of cell towers fighting over a given phone.
At my parents' house, Verizon phones get zero coverage, despite three towers being within range.
The same arguments were abound when Intel introduced HT. And Microsoft didn't touch the licensing.
My guess is XP Home will continue to only use one logical processor, while XP Pro will use two. (Now, whether the "second" logical processor is HT on the first core, or primary use on the second core, remains to be seen.)
That's funny, 'cause before the McDonalds down the street from where I worked turned off the order-confirmation screen a few months ago, it'd been showing a Windows GPF error for several weeks...
Requires stability. Indeed.
I told them to power the machine off and back on, and they said they'd tried that. They were still waiting for someone from McD's IT department to come down and fix the box.
To cook the occupants. Not nice. But stil effective.
No...that'd be the two universes mentioned in the TOS episode "What of Lazarus"
What of Lazarus...and what of Lazarus...
I didn't realize Douglas Adams was posting from beyond the grave...
...it'd become an RFC at some point.
How about a Linux release?
Wonderful publicity. Way to encourage them.
Click on "and more" ...
Google attempts to group related articles.
...Valve was waiting for the arrests before releasing the game.
Another poster mentioned it did, but I can't find a reference.
Assuming the ruling goes against them, you know they'll appeal. Even if it doesn't, IBM will.
I don't think they have the money to follow through all the appeals.
Yeah...you know they can't work with an open-sourced engine. :)
Won't be necessary. SCO is a corporation, which means its officers are protected from legal action stemming from the behavior of the company..
IANAL.
I'd be more worried about some sort of action against Groklaw.
The problem I see is the difficulty in producing a machine that can not only reproduce itself, but do something useful as well.
I think a much more likely scenario would be producing a solution consisting of two machine species. One species does useful work. The other species makes copies of the first species.
The two species would depend on different keyed source chemicals. If you wanted to increase the rate at which you build load-bearing bots, you increase the availability of one chemical. If you wanted to increase the rate at which you were doing useful work, you increase the availability of the other chemical.
Mix corn starch and water. There's some interesting goo for you to play with.
Hit it hard with your fingers, and they'll stop at the surface. But you can slowly sink your fingers in without much fuss.
If this was attempted in America, I wonder if it would be considered as allowing "unreasonable search and seizure."
If you run Linux, I'd wager that the source code was available for almost all of the software on your machine. Meaning it can be compiled with optimizations for your CPU.
With Via's CPUs, some functionality, like encryption and random number generation, are supported by the kernel, which then makes the functionality available to userland in ways that don't require the software to explicitly support Via's CPU.
I was referring primarily to AMD versus Intel CPUs. Not only do they support different instruction set extensions, you need to provide your instructions at different intervals and orders in order to get the most out of the CPU.
Not to mention x86-64...
DVD install media only holds 4.7GB, there are several distros which are already approaching (or have already passed) this if you add up all the CDs...
All of "testing" Debian requires 13 CDs, or 3 DVDs. I've got the 13-CD set.
Microsoft has the advantage that most of their end-users got Windows on their machines pre-installed. That means they can include a lot more data in their OS, which means they can add code optimized for the popular processors.
I wouldn't be surprised to see similar efforts in the Linux community as, for example, DVD install media becomes more common, but there's still one problem. gcc doesn't make optimization for multiple systems easy. mplayer does it, but that's all I know of.
I wanted to put rudimentary FEA in the Quake engine...imagine causing the opposing team's fort to collapse.
Unfortunately, I lacked the programming skills--and still lack the mathematics skills--to do it.