The way I see it, a PPC card would only have a limited amount of use. It would be damned expensive, for one. Not only is there cost of the CPU, but you're going to need a full-fledged chipset and a good hunk of memory, if you don't intend to limit your memory throughput with the PCI bus.
Or you could build it as an AGP card. That'll murder your graphics capability.
Finally, you have to cool the thing. That much hardware is going to require active cooling, especially considering it's not the only CPU in the machine.
Re:IT/Clustering made cheap
on
ClusterKnoppix
·
· Score: 1
Yup. OpenMOSIX doesn't do anything to interfere with userland activities.
IT/Clustering made cheap
on
ClusterKnoppix
·
· Score: 2, Interesting
Linux Magazine just did a three-article bonanza on how Beowulf clusters came about, and the costs and issues involved with running one.
This could blow all of that away; Just insert the CD in all of the machines in your office, and let fly. Air conditioning? Already accounted for. Power consumption? Not much more than usual. Floorspace? Just a little under everyone's desk.
What I'd like to see would be companies switching over to all-Linux or mostly-Linux shops, running all their machines as an OpenMosix cluster. They could sell off their spare CPU cycles, quite easily. Ironically, IT's never been so cheap.
You would need a platform one-ninth the performance of his Athlon for each cell phone.
The easiest way to do that would probably be to build the capability into the phone, which means new phones for most people, firmware upgrades for everyone else.
I can't imagine they'd want to buy an Athlon box for every nine cell phones they serve. But then, the parent didn't mention the bitrate he was using.
Can you give us a reference? I've never heard this before...
Remember that TCP/IP came before OSI, and became successful under a different system. Does that mean it has to be thrown away when a new rating system comes along? Millions of people love The Matrix, but that doesn't mean they won't admit that Gone With The Wind is a classic.
My grandmother is terrified of all arachnids and reptiles. During the fifties, she shot a snake that was in the road next to her mailbox.
Due to various physical difficulties, she and my grandfather don't hug each other often, so I'd much rather her wear one of these than have her carry a gun. (She doesn't have the physical dexterity to evade a grappling attack, or to keep control of her revolver when at a disadvantage.)
I'm moderately ignorant on the subject, and NVidia always seemed to be the leader to me. All I ever hear about is how NVidia's still ahead. (Though the recent discovery that they rigged their scores for at least one benchmark certainly throws a monkey wrench into things)
On Slashdot, when it comes to video cards, I usually only read the headlines and editor blurbs. Those have usually been pro-NVidia.
I'd like to put forward that the majority of people are at least as ignorant as I am, and that whoever majority opinion favors, that company is probably the recipient of the most sales.
Orson Swindle, one of two Federal Trade Commission members to testify, urged the industry to develop technology that could stop mail from any source except those pre-approved by users.
"There is a basic need for consumers to be free of any unwanted e-mail," Swindle said.
There's something wrong with an FCC commisioner named "swindle"...
Not punishment in terms of jailtime, certainly. I'm thinking along the lines of probation or fines, such that the parent will take steps to keep their children from misbehaving again.
It's true that many parents of problem children are doing the best they know how to raise them; the need is therefore shown that they need to learn more effective means of parenting.
I'm not saying we force specific techniques down their throats; that's a form of government control I'd rather not take, and it would tend to eliminate diversity in problem-solving in children. I'm saying they should be encouraged to go out and find better techniques on their own.
The biggest benefit is there would be a popular push for research into parenting techniques.
...of legislators taking responsibility out of the parents' hands.
There ought to be punishment for parents whose kids break laws. That should encourage parents to raise their kids to respect rules.
Most of the "punks" I met in high school had parents who didn't care what they did. In fact, a few of them would have their parents bail them out at school whenever they could.
It's a cheap way to get one to take apart, though. We've all heard about the principles they use to steer, but I don't think anyone's ever seen a schematic.
In order to ensure proper coding standards, you'd have to have a defined minimum level that applications had to reach before they could be used, as en example, on the Internet.
Who's going to define that minimum level? An elected body? Government appointees?
Currently, that minimum level is defined by practicality. People only use broken software if it is practical for them to do so. You'd have to take after Microsoft if you want to force other people to adhere to your personal standards.
What you're advocating is catagorization based on phenotype. That's a debate that's been going on for a long time.
The problem with that system is how much alike do two creatures have to be in order to be in the same family? If you're not specific enough, you might end up placing hummingbirds and flies in the same family because they both have superb control over their flight characteristics, and they both like sugarwater.
If you're too specific, you'll start separating (as an example from history) people of African descent from people of European descent.
Classifying species based on genotype allows us to trace nature's evolutionary path, and understand biological history accordingly.
I'm not going to make any grand comparisons between your brain and, well, anything else.
He has a point; there is a great deal of the human genome that seems to be useless, in that we don't know what it does. There have been experiments with bacteria to try to find out how much of the DNA is unneccessary. When the presumed useless material is removed from a bacterium, however, the cell dies.
We're also another version of an already successful line of mammals, or is that going unreasonably far, in your definition of reason?
Having different genuses is one of biology's ways of quantifying the differences between types of organisms; indeed, it helps define what constitutes a "type."
It depends on the application. The V240's CPUs may be 1GHZ, but they have twice the cache of Intel's Xeons. I'm willing to bet that the cache difference will have a large effect on databases and image processing.
Slashdot Mirror
The way I see it, a PPC card would only have a limited amount of use. It would be damned expensive, for one. Not only is there cost of the CPU, but you're going to need a full-fledged chipset and a good hunk of memory, if you don't intend to limit your memory throughput with the PCI bus.
Or you could build it as an AGP card. That'll murder your graphics capability.
Finally, you have to cool the thing. That much hardware is going to require active cooling, especially considering it's not the only CPU in the machine.
Yup. OpenMOSIX doesn't do anything to interfere with userland activities.
Linux Magazine just did a three-article bonanza on how Beowulf clusters came about, and the costs and issues involved with running one.
This could blow all of that away; Just insert the CD in all of the machines in your office, and let fly. Air conditioning? Already accounted for. Power consumption? Not much more than usual. Floorspace? Just a little under everyone's desk.
What I'd like to see would be companies switching over to all-Linux or mostly-Linux shops, running all their machines as an OpenMosix cluster. They could sell off their spare CPU cycles, quite easily. Ironically, IT's never been so cheap.
I can see the next press release now:
"They were using the equivalent of 16 PCs...er...nevermind."
This will be an interesting study? How useful will multiple independant bittorrents be?
If you have multiple Torrent servers/trackers, won't you be limiting the available bandwitdth?
Um...this is probably obvious, but why doesn't someone set up a Torrent?
Hmm. Liquid War comes to mind. I certainly like Heroes (even if it still vaguely resembles Tron.)
I guess you need to go out and destroy the Internet, then, huh?
You would need a platform one-ninth the performance of his Athlon for each cell phone.
The easiest way to do that would probably be to build the capability into the phone, which means new phones for most people, firmware upgrades for everyone else.
I can't imagine they'd want to buy an Athlon box for every nine cell phones they serve. But then, the parent didn't mention the bitrate he was using.
Can you give us a reference? I've never heard this before...
Remember that TCP/IP came before OSI, and became successful under a different system. Does that mean it has to be thrown away when a new rating system comes along? Millions of people love The Matrix, but that doesn't mean they won't admit that Gone With The Wind is a classic.
My grandmother is terrified of all arachnids and reptiles. During the fifties, she shot a snake that was in the road next to her mailbox.
Due to various physical difficulties, she and my grandfather don't hug each other often, so I'd much rather her wear one of these than have her carry a gun. (She doesn't have the physical dexterity to evade a grappling attack, or to keep control of her revolver when at a disadvantage.)
I'm moderately ignorant on the subject, and NVidia always seemed to be the leader to me. All I ever hear about is how NVidia's still ahead. (Though the recent discovery that they rigged their scores for at least one benchmark certainly throws a monkey wrench into things)
On Slashdot, when it comes to video cards, I usually only read the headlines and editor blurbs. Those have usually been pro-NVidia.
I'd like to put forward that the majority of people are at least as ignorant as I am, and that whoever majority opinion favors, that company is probably the recipient of the most sales.
Orson Swindle, one of two Federal Trade Commission members to testify, urged the industry to develop technology that could stop mail from any source except those pre-approved by users.
"There is a basic need for consumers to be free of any unwanted e-mail," Swindle said.
There's something wrong with an FCC commisioner named "swindle"...
Not punishment in terms of jailtime, certainly. I'm thinking along the lines of probation or fines, such that the parent will take steps to keep their children from misbehaving again.
It's true that many parents of problem children are doing the best they know how to raise them; the need is therefore shown that they need to learn more effective means of parenting.
I'm not saying we force specific techniques down their throats; that's a form of government control I'd rather not take, and it would tend to eliminate diversity in problem-solving in children. I'm saying they should be encouraged to go out and find better techniques on their own.
The biggest benefit is there would be a popular push for research into parenting techniques.
...of legislators taking responsibility out of the parents' hands.
There ought to be punishment for parents whose kids break laws. That should encourage parents to raise their kids to respect rules.
Most of the "punks" I met in high school had parents who didn't care what they did. In fact, a few of them would have their parents bail them out at school whenever they could.
It's a cheap way to get one to take apart, though. We've all heard about the principles they use to steer, but I don't think anyone's ever seen a schematic.
Are you going to compare every line of code in both codebases?
There is some level of similar functionality, and that functionality is probably what SCO is going to claim is theirs.
Imagine some old-timer claiming the rights to heirarchial filesystems, and you may see some similarities.
In order to ensure proper coding standards, you'd have to have a defined minimum level that applications had to reach before they could be used, as en example, on the Internet.
Who's going to define that minimum level? An elected body? Government appointees?
Currently, that minimum level is defined by practicality. People only use broken software if it is practical for them to do so. You'd have to take after Microsoft if you want to force other people to adhere to your personal standards.
What you're advocating is catagorization based on phenotype. That's a debate that's been going on for a long time.
The problem with that system is how much alike do two creatures have to be in order to be in the same family? If you're not specific enough, you might end up placing hummingbirds and flies in the same family because they both have superb control over their flight characteristics, and they both like sugarwater.
If you're too specific, you'll start separating (as an example from history) people of African descent from people of European descent.
Classifying species based on genotype allows us to trace nature's evolutionary path, and understand biological history accordingly.
I'm not going to make any grand comparisons between your brain and, well, anything else.
He has a point; there is a great deal of the human genome that seems to be useless, in that we don't know what it does. There have been experiments with bacteria to try to find out how much of the DNA is unneccessary. When the presumed useless material is removed from a bacterium, however, the cell dies.
We're also another version of an already successful line of mammals, or is that going unreasonably far, in your definition of reason?
Having different genuses is one of biology's ways of quantifying the differences between types of organisms; indeed, it helps define what constitutes a "type."
Adding a whole bunch of taxpayers to the pool would mean a massive increase in revenue, without having any effect on the lower per-person tax rate.
The question is, while they evolved, did their genome evolve in a direction similar enough to our own to warrant including them in the genus homo?
To me, it just seems to be an argument debating the differences between genotype and phenotype.
If 99% of the important DNA is identical, then probability implies that 99% of the rest of the DNA is also identical.
It depends on the application. The V240's CPUs may be 1GHZ, but they have twice the cache of Intel's Xeons. I'm willing to bet that the cache difference will have a large effect on databases and image processing.