The NetBurst architecture in the P4 is legendary for it's terrible performance per clock.
Authors of applications intended for said architecture were probably aware of its terrible IPC. This means the applications will run fine on Atom, which has IPC that's similarly terrible but for a different reason.
I wouldn't count on it. Atom also has a lame ass pipeline.
While AMD focused in increasing performance per clock with the Athlon, Intel intended to increase performance through higher clock speeds.
Higher clock and longer pipeline. The latter was the real killer. Fine for straight up linear code, but not so much if the pipeline stalls multiple times per microsecond.
As I understand it, an Atom CPU roughly matches the performance of a Pentium 4 CPU clock for clock. This means an old application that runs well on a P4 will also run well on an Atom.
Sad to say, I bought Atom based notebooks a total of three times, having failed to learn my lesson the first time. Disappointed every time despite claims the the effect that it's better this time, honest! The truth: low performing, desk melting, battery chewing junk that I regret owning. Thanks for the effort Intel, but there will be no fourth Atom attempt from me.
On the other hand, every ARM-based device I have (about 8 now) is a pleasure to use.
(...but I have a Pentium 4 based shoebox that is still going strong, go figure)
(maybe do more public relations and github, if you don't)
This. Being able to point to your work in a public repository for a project actually doing something useful speaks volumes and cuts down the effort needed to evaluate you enormously. Establish some open source cred and the world is your oyster.
Is Linux successful? Debatable. It has success in limited uses, but has never grown beyond these uses. Android is a Linux fork. It's in a lot of pockets and a lot of homes.
Agreed, Linus is no one trick pony. But why do you suppose he named it git? Could it have anything to do with some stupid git virtually tearing the kernel community in two by forcing a proprietary product into the Linux kernel tool chain?
Yes, we're still hating on Gnome because it is technically poor and long outlived its intended purpose, which was to force QT to go to a free license. Nowadays Gnome's sole purpose is to be a big stick used by Redhat to beat up other, more deserving and technically better projects, the poor long suffering user being caught in the middle.
Sarah is far from the only talented kernel dev to leave because it just stops being fun. And competes with other projects with a healthier culture. Just in case anybody claims that the usual toxic waste is actually the best and fastest way to develop the kernel, consider that Linux still has crappy real time response, all you need to do is start a buch of glgears, then copy a big file to see that. And simple file copies still tend to page out the application code you're running. And the kernel is way bloatier than necessary because the kind of people who enjoy cleaning that kind of crap up tend to get disparaged or ignored, and just walk away.
If there is no demand then there is no replacement, hence no replacement cost. If there is high demand then replacement cost will be higher because the price of replacement will go up. Law of supply and demand.
It is inevitable that some organism has inherited so much of its genome by reverse transcriptase etc as to muddy the question of exactly what it descended from. Then artificial species will start showing up and it alls gets even messier than it is now. I wonder if it could eventually have cycles.
The Skype servers all run Linux. I guess Microsoft finally got burned enough times trying to make IIS do what Linux does, to stay well enough away from that. But they did reconfigure the server topology to make it easier to spy on corporate traffic for example, or anything else they feel like.
The text client does have one cool feature... you can correct mistakes in your last posted line using sed syntax. How geeky is that?
Of course, that just shows where Skype came from. I'm sure Microsoft would have come up with a much clunkier way of doing the same thing, or just not do it.
Slashdot Mirror
The NetBurst architecture in the P4 is legendary for it's terrible performance per clock.
Authors of applications intended for said architecture were probably aware of its terrible IPC. This means the applications will run fine on Atom, which has IPC that's similarly terrible but for a different reason.
I wouldn't count on it. Atom also has a lame ass pipeline.
While AMD focused in increasing performance per clock with the Athlon, Intel intended to increase performance through higher clock speeds.
Higher clock and longer pipeline. The latter was the real killer. Fine for straight up linear code, but not so much if the pipeline stalls multiple times per microsecond.
As I understand it, an Atom CPU roughly matches the performance of a Pentium 4 CPU clock for clock. This means an old application that runs well on a P4 will also run well on an Atom.
Sad to say, I bought Atom based notebooks a total of three times, having failed to learn my lesson the first time. Disappointed every time despite claims the the effect that it's better this time, honest! The truth: low performing, desk melting, battery chewing junk that I regret owning. Thanks for the effort Intel, but there will be no fourth Atom attempt from me.
On the other hand, every ARM-based device I have (about 8 now) is a pleasure to use.
(...but I have a Pentium 4 based shoebox that is still going strong, go figure)
The missing sd slot is the only thing that keeps me from buying this phone.
Analog instruments in a typical aircraft "six pack" are quite useful when in the air but mostly useless if floating on water...
The altimeter will double as a barometer.
(maybe do more public relations and github, if you don't)
This. Being able to point to your work in a public repository for a project actually doing something useful speaks volumes and cuts down the effort needed to evaluate you enormously. Establish some open source cred and the world is your oyster.
Moore's law is about transistor density, not the number of cores or their performance.
Actually, it's about transistor count.
I know plenty of geeks who are properly socialized and capable of normative human relations. If you are not one, perhaps you should become one.
Is Linux successful? Debatable. It has success in limited uses, but has never grown beyond these uses.
Android is a Linux fork. It's in a lot of pockets and a lot of homes.
Android is not a Linux fork, Android is Linux.
You didn't mention the embedded devices. How many Linux computers do you think the average person has running in their home, even a Linux hater?
People use it in those places because it's free, not because it's particularly good...
I just love you comedians, a good belly laugh once a day is just the thing to help me live longer.
Agreed, Linus is no one trick pony. But why do you suppose he named it git? Could it have anything to do with some stupid git virtually tearing the kernel community in two by forcing a proprietary product into the Linux kernel tool chain?
Yes, we're still hating on Gnome because it is technically poor and long outlived its intended purpose, which was to force QT to go to a free license. Nowadays Gnome's sole purpose is to be a big stick used by Redhat to beat up other, more deserving and technically better projects, the poor long suffering user being caught in the middle.
Sarah is far from the only talented kernel dev to leave because it just stops being fun. And competes with other projects with a healthier culture. Just in case anybody claims that the usual toxic waste is actually the best and fastest way to develop the kernel, consider that Linux still has crappy real time response, all you need to do is start a buch of glgears, then copy a big file to see that. And simple file copies still tend to page out the application code you're running. And the kernel is way bloatier than necessary because the kind of people who enjoy cleaning that kind of crap up tend to get disparaged or ignored, and just walk away.
If there is no demand then there is no replacement, hence no replacement cost. If there is high demand then replacement cost will be higher because the price of replacement will go up. Law of supply and demand.
Bearing in mind that several trillion dollars of industry now depend on it.
Why bear that in mind? It has nothing to do with the replacement cost.
Law of supply and demand.
I can't read the paper since I need to fill out some fscking form from, and that's not happening.
That bit of PHB genius demonstrates adequately how disconnected from the community the Linux Foundation really is.
Does the Linux Foundation own either of those two things?
The Linux Foundation does not "own" any of its projects.
The GCC compiler, for instance, is widely used, and it's disappearance would put a large hole in the software world. Gnu Hurd, on the other hand....
Neither is a Linux Foundation project.
What about Linux kernel? Bearing in mind that several trillion dollars of industry now depend on it.
It is inevitable that some organism has inherited so much of its genome by reverse transcriptase etc as to muddy the question of exactly what it descended from. Then artificial species will start showing up and it alls gets even messier than it is now. I wonder if it could eventually have cycles.
It's like making an English dictionary, except way worse.
HTML by itself is not Turing complete. You need CSS for that.
The Skype servers all run Linux. I guess Microsoft finally got burned enough times trying to make IIS do what Linux does, to stay well enough away from that. But they did reconfigure the server topology to make it easier to spy on corporate traffic for example, or anything else they feel like.
The text client does have one cool feature... you can correct mistakes in your last posted line using sed syntax. How geeky is that?
Of course, that just shows where Skype came from. I'm sure Microsoft would have come up with a much clunkier way of doing the same thing, or just not do it.