Microsoft's way of being multiplatform is to simply kill all other platforms. This would certainly go hand and hand with the recent 2 Billion blunder. Go read Cringly's latest and you'll see some very interesting similarities between what's happening and what he's predicted.
Sadly, Intel isn't Apple. Intel's never been about marketing their product based on how well it's completed a benchmark; just how fast it can clock matters to most consumers because it's a big, flashy number, and big flashy numbers are quiet distracting.
Apple got it right by using Benchmarks to sell their product, even if the benchmarks are strange and deceptive. Hey, lying, cheating, and stealing are what got Microsoft to the top, everyone's gotta play a little dirty.
And yes, buying a PC should be an emotional experience, as well as a scientific one. PC's, like Cars, are where us humans now spend a great deal of our lives. (nerds/geeks especially). We need to have some level of attachment to the machine we're using just so that it doesn't drive us mad.
Not to mention increasing the higher leak current, the possibility to increase the clockrate to 5GHz, and higher power consumption. Oh, the 5GHz was an estimate of Intel. It is their current target for the end of the year. So, no Dothan on the desktop. 5GHz may be their target, and they may hit/miss it with Prescott, but the fact is, what Prescott is for, has changed. They know as well as we do that this is a totally unacceptable chip for the desktop (except for the extreme high end, gamers, case modders), and likewise, they've made a motherboard format for those communities (Pico and Micro BTX). Prescott will most likely be dismissed internally as a mistake, passed into the Xeon line, where Tejas will pick up later. So, they're still hitting 5GHz, but it's a big whoop, as their desktop processor Dothon will be hitting 6000+ (and only running roughly 3.6Ghz, if my predictions on the scalability of Banias and Dothan hold out)
The 3.2GHz Prescott consumes even a fair amount more energy than its 3.2Ghz predecessor.
If I were Intel, I'd (falsely) argue that the power it uses extra makes it a more reliable chip for a server, and therefore should be there. Companies are going to hate that it's a thermal hogg, but they'll buy because it's Intel, and they've got the best reputation.
No, it is because the mainboards and psu can't deliver the 100A those devices would require. And it is quite a problem to dissipate the heat of such a thing. Remember the new mainboard-layout which shoud cope with that thing? Also an idea of Intel.
In fact, I believe they not only could, but are, delivering more than enough power, although your old 200 Watt would have to say goodnight. Most PC's built by Dell ship with a proprietary atx-like PSU running a measly 220 Watts or so, on the P4 line none-the-less. The power subsystem's there, but using it is more of a risk than anyone wants to take, no?
Read the title for your reduced power consumption chips, which should be hitting the desktop within a few months or so. Banias is just wiping the floors with any competitors battery-life and speed wise, and their greatest competitor is actually themselves; those god awful Celeron notebooks with 30 minute battery lifes. But what's cheap usually outsells what's new.
I fully believe that Pentium V (Pentium 5, whichever they choose to call it), will be Dothan, introducing to the desktop for the first time a power-saving logic. Not only does this make sense for quieter, smaller computing (two of the biggest buzzfactors on the market right now; those micro cases and motherboards are selling like wildfire), it makes for cheaper, faster computing. I believe that the cluncky Pentium 4M will be dropped altogether, and the Pentium 4 technology (Tejas, the last NetBurst Archetecture chip I know of) will be integrated into the Xeon line to run head on verses the AMD64 chips (hince, the reason they're adding in the x86-64 extensions to that processor).
AMD may be using numbers like 2300+ to describe the speed, but in the end, when a person goes to their local Walmart, or Dell.com, or whereever they go to buy their next PC, they're only going to look as far as "hmm, 2300+ is bigger than 2200+". They're not even going to know what the actual speed difference is, because they don't care to know, just as long as what they're getting is faster. GHz IMO is at least a little more honest when it comes to Intel Processors because the IPC (instructions per clock) shouldnt change all that much from a 2.0GHz CPU to a 2.2 GHz CPU, whereas the instructions per clock on a 2600+ CPU can be drastically different from that of a 2700+ (in fact, it can be a whole different core). Also true for the Pentium 4's as well. Damn, we just need a standard... someone, anyone, PLEASE!?
Okay, I thought I'd sit back and moderate on this one, but I'm already tired of reading the garbage.
Ever wonder why Intel's not been cranking out Prescott cored processors that run even faster/hotter? Is it because they couldn't just bolt a jet engine and a copper block to the thing and ship it? No. It's because they're shifting their attention (once again).
AMD fanboys listen up: Yeah, you guys are winning the strongarm race right now. You've got the faster middle-class processor (upper end desktop/lower- to medium-end server) and Intel knows this quite well. They could scale Prescott very quickly up, but so would come heat, and therefore energy prices.
Now, lets look at other moves Intel's made lately. They've announced they're going to a PR-rating for selling processors. What sense does this make if they're just going to ramp up their processors even faster clockwise? Why do they need to compare anything except clocks? Well, because AMD is wiping the floor with them, that's why.
90nm technology has also been undergoing perfection with Prescott, meaning lower voltages, higher yields, less wasted silicion on the wafer.
Both of these things bring us to the sucecssor to Banias, Dothan. Extremely large cache, 90nm technology, extremely fast CPU. Not only will this be one of the most (energy effecient/clock effecient) chips ever made, it most likely will be the next desktop processor. But, here's the kicker, for them to be able to do this, they need to take a short pause from ramping up their current technology's speed, and moving Dothan over to a bigger mass production line. This is why we find Intel pretty silent right now, and most likely the same with AMD (anticipation; AMD's a very reactionary company).
So, I'm very sure that this is one of the top priorities sitting on the desks of Intel Engineers, I applaud them for taking every step towards a cleaner environment while making my newest tech gadgets.
well, maybe he's just looking for a better tool, maybe ethereal, tcpdump, and etherpeak don't offer a specific feature he's looking for (although, i have no clue what it might be, looking at UDP packet?)
The registry is actually quite a good idea, with a horrible implementation. Basically, it shouldn't be a single file database-style solution. This adds confusion that isn't needed. The other reason it was poorly implemented was because they continued to use INF files, instead of depreciating the entire subsystem, for removal in windows 98 (95 was the first instance of the "system registry" as we've came to know it).
That's probably the whole reason Windows is in as bad of shape as it is; lack of standards, or downright bastardization, reinvention, or some combination of the two.
In theory, they could fix the registry, and I'm working on a similar Linux project right now. The idea of having directories redefineable (think: XML file; sorta what microsoft did with Desktop.ini) based on content, and virtual directories that have any content that fits a certain layout (think: Virtual MP3 Directory, Virtual Movie Directory, Virtual Config file Directory). Defining a virtual conf dir would allow you a one stop location to fix whatever system problem you wanted, and you could even make a purdy little gtk2 or qt3 app made specifically for editing these files (although, it would end up becoming something like G-Edit with a Sidebar).
Of course, this can be implemented by microsoft too, since their new WinFS will be an SQL-based file system. Using a Redefineable directory like "C:\Windows\Configuration Files\" could eventually replace the "registry", and help Windows become a whole lot more secure (by setting appropriate permissions on this directory).
Usually, ideas are very sound and awesome pieces of work. It's the implementations that are the bitch...
The minute Microsoft pushes their crap into a bios chip is the minute I move to Apple computers. Mac's are better anyways [/flameoff]
LinuxBIOS really isn't for everyone. It's not as hardware agnostic as a typical harddrive based bootloader is, simply because a bios chip has a tiny amount of eeprom (256k). While this is suffiecient for an optimized system (cluster node, web server, invariant hardware, etc), it's really not good for a home user who'd love to throw his/her new radeon in and not have to insmod their driver everytime they bootup. It's also hell to recompile the kernel, as one false move and your bios is history.
One thing that may be beautiful: flash ram being used instead of current bios archetecture. Flash ram is cheap, and spacious, and can hold quite a large kernel-image. But this gets into more of what XIP is trying to do....
I understand your concerns of Linux becoming Microsoft-ish, and locking in people, but come on, it's just a BIOS image. Don't want it, go to the vendor's website, download the newest one, write it on. Of course this is more dangerous, but for that same reason, I don't believe we'll be seeing a huge push for LinuxBIOS any time soon.
Errr... I really need to re-read what I write. It's a different THEORY, but similar execution.. eXecution In Place is speed via giving the computer EXACTLY what it needs to run, whereas LinuxBIOS is simply skiping the step of harddrive start. Completely different in the fact they're starting from opposite ends of the same problem, but slight in the fact that they've reached an almost common ground.
Fuck grub, this is moving the whole kernel into flash-ram. This is to save the step of having to connect to the hard drive until when it's actually needed to run software, which can be done WHILE the system is booting the kernel, instead of before-hand. And as we know, in the computer world, multitasking is a good way of speeding up things.
Well, for starters if my understanding of eXecute In Place is holding, it's completely different. The ROM chip holds a compressed kernel, that's decompressed on the fly and used to start the system. The only thing this skips is the need to do a startup seek for harddrives, and that whole step of initalizing slow hardware (ie, Hard Disk platters). eXecute In Place is intented to simply run the program natively, ie uncompressed kernel image sitting in ram, so it can just be thrown straight into the processor and executed. The difference is slight, but none-the-less important.
Not from how I understand it. This is actually saving a Kernel-Image onto the bios chip itself, therefore booting straight to the kernel from the chip, basically skipping the step of "BIOS tells HD to find MBR, then read the data from MBR to ram, execute MBR instruction". Or at least, that's how I understand it?
But once again, I can reaffirm to you that I'm choosing my profession because I love it.. (hate perl, but that's a whole 'nother story... damned syntatically obsfusecated language...real men use python;)
Sure, as you all can call me young, you can cite my inexperience, but you can't cite my intelegence, nor can you cite my ambishions nor my abilities. In 4 years, 6 years, 10 years from now, I'll be happy one way or another, because in one way or another, there will always be a niche that needs to be filled where outsourcing is just not an option, even if this does mean I'm not doing super cool innovative work, and not something more than a site manager or something like that. The fact is, there will always be jobs in computers, finding them may be hard, but if you love your work as much as I do, than you have the motivation required.
Oh, and about the mortgage, 2.5 kids, an ex-wife, SUV thing: I'm a firm believer in using condoms, won't get married for a long time *trust me, I've had my share of bad relationships at this age to know that no Sure Thing (tm) is worth the impulse.., an enviromentalist (as much as I can be...), will drive a Gas Electric car as soon as I can afford to purchase one;), hate credit cards (too much power corrupts), oh, and did I mention I loved my work with computers?
Sorry your life crapped out dude, but I refuse to run mine like that. And I'm worried too about the shift, it's not that I'm not worried, it's that times are changing. I'm going to do computers because I love computers, not because of money, not because of anything else. I love the ability to extend the mind into a chip, to do work at a rate unfathomable by most humans, and the ability to improve our lives that exists within them. These are my aspirations, my goals while working with computers, getting rich is about Null on that list.
One last note: I didn't say live with it, I said "adapt". Change, become something new, be dynamic, force change. Innovate, make your superiors notice you, do work that truly becons being done, not work that's painful and agonizing. Don't mourn your life, LIVE IT!!!!!
I'm to be a freshman in the fall as well, and I'm still going to Major in Computer Engineering, but only because I want to do something that I truely love, and really don't care about the pay. The truth is these jobs are going over seas, and they're moving quickly, but as we've seen with most of these job fluctuations, they tend to be short term (think: NAFTA and the like...), and they tend to stablize themselves quickly. Worst comes to worse, I'll move to India;)
I think the biggest reason today that jobs are shifting overseas is simply the costs of running a redundant business. Very few companies are actually innovating these days, and those that are, do their work in the good ol' USA because of strong patenting laws (yes, too strong, we know..) and the like. Those same companies are offsetting the price of innovation by reducing the cost of tech support, sending it offshores to cheaper labor. I think the best way to get out of this is simply a change in buisness model; too many buisnesses are worried about the upfront costs as compared to the long-range profits to be gained, and are getting downright greedy and stingy when it comes to money...
Basically, the economic structure of America is changing. Don't like it? Move. Or stay here and adapt.
Slashdot Mirror
Microsoft's way of being multiplatform is to simply kill all other platforms. This would certainly go hand and hand with the recent 2 Billion blunder. Go read Cringly's latest and you'll see some very interesting similarities between what's happening and what he's predicted.
Uhhhh...
Couldn't have said it any better myself. Thanks.
Sadly, Intel isn't Apple. Intel's never been about marketing their product based on how well it's completed a benchmark; just how fast it can clock matters to most consumers because it's a big, flashy number, and big flashy numbers are quiet distracting.
Apple got it right by using Benchmarks to sell their product, even if the benchmarks are strange and deceptive. Hey, lying, cheating, and stealing are what got Microsoft to the top, everyone's gotta play a little dirty.
And yes, buying a PC should be an emotional experience, as well as a scientific one. PC's, like Cars, are where us humans now spend a great deal of our lives. (nerds/geeks especially). We need to have some level of attachment to the machine we're using just so that it doesn't drive us mad.
http://www.wired.com/news/business/0,1367,62851,00 .html?tw=wn_tophead_1
Not to mention increasing the higher leak current, the possibility to increase the clockrate to 5GHz, and higher power consumption. Oh, the 5GHz was an estimate of Intel. It is their current target for the end of the year. So, no Dothan on the desktop.
5GHz may be their target, and they may hit/miss it with Prescott, but the fact is, what Prescott is for, has changed. They know as well as we do that this is a totally unacceptable chip for the desktop (except for the extreme high end, gamers, case modders), and likewise, they've made a motherboard format for those communities (Pico and Micro BTX). Prescott will most likely be dismissed internally as a mistake, passed into the Xeon line, where Tejas will pick up later. So, they're still hitting 5GHz, but it's a big whoop, as their desktop processor Dothon will be hitting 6000+ (and only running roughly 3.6Ghz, if my predictions on the scalability of Banias and Dothan hold out) The 3.2GHz Prescott consumes even a fair amount more energy than its 3.2Ghz predecessor.
If I were Intel, I'd (falsely) argue that the power it uses extra makes it a more reliable chip for a server, and therefore should be there. Companies are going to hate that it's a thermal hogg, but they'll buy because it's Intel, and they've got the best reputation. No, it is because the mainboards and psu can't deliver the 100A those devices would require. And it is quite a problem to dissipate the heat of such a thing. Remember the new mainboard-layout which shoud cope with that thing? Also an idea of Intel.
In fact, I believe they not only could, but are, delivering more than enough power, although your old 200 Watt would have to say goodnight. Most PC's built by Dell ship with a proprietary atx-like PSU running a measly 220 Watts or so, on the P4 line none-the-less. The power subsystem's there, but using it is more of a risk than anyone wants to take, no?
Read the title for your reduced power consumption chips, which should be hitting the desktop within a few months or so. Banias is just wiping the floors with any competitors battery-life and speed wise, and their greatest competitor is actually themselves; those god awful Celeron notebooks with 30 minute battery lifes. But what's cheap usually outsells what's new.
I fully believe that Pentium V (Pentium 5, whichever they choose to call it), will be Dothan, introducing to the desktop for the first time a power-saving logic. Not only does this make sense for quieter, smaller computing (two of the biggest buzzfactors on the market right now; those micro cases and motherboards are selling like wildfire), it makes for cheaper, faster computing. I believe that the cluncky Pentium 4M will be dropped altogether, and the Pentium 4 technology (Tejas, the last NetBurst Archetecture chip I know of) will be integrated into the Xeon line to run head on verses the AMD64 chips (hince, the reason they're adding in the x86-64 extensions to that processor).
Long Live P6
AMD may be using numbers like 2300+ to describe the speed, but in the end, when a person goes to their local Walmart, or Dell.com, or whereever they go to buy their next PC, they're only going to look as far as "hmm, 2300+ is bigger than 2200+". They're not even going to know what the actual speed difference is, because they don't care to know, just as long as what they're getting is faster. GHz IMO is at least a little more honest when it comes to Intel Processors because the IPC (instructions per clock) shouldnt change all that much from a 2.0GHz CPU to a 2.2 GHz CPU, whereas the instructions per clock on a 2600+ CPU can be drastically different from that of a 2700+ (in fact, it can be a whole different core). Also true for the Pentium 4's as well. Damn, we just need a standard... someone, anyone, PLEASE!?
Okay, I thought I'd sit back and moderate on this one, but I'm already tired of reading the garbage.
Ever wonder why Intel's not been cranking out Prescott cored processors that run even faster/hotter? Is it because they couldn't just bolt a jet engine and a copper block to the thing and ship it? No. It's because they're shifting their attention (once again).
AMD fanboys listen up: Yeah, you guys are winning the strongarm race right now. You've got the faster middle-class processor (upper end desktop/lower- to medium-end server) and Intel knows this quite well. They could scale Prescott very quickly up, but so would come heat, and therefore energy prices.
Now, lets look at other moves Intel's made lately. They've announced they're going to a PR-rating for selling processors. What sense does this make if they're just going to ramp up their processors even faster clockwise? Why do they need to compare anything except clocks? Well, because AMD is wiping the floor with them, that's why.
90nm technology has also been undergoing perfection with Prescott, meaning lower voltages, higher yields, less wasted silicion on the wafer.
Both of these things bring us to the sucecssor to Banias, Dothan. Extremely large cache, 90nm technology, extremely fast CPU. Not only will this be one of the most (energy effecient/clock effecient) chips ever made, it most likely will be the next desktop processor. But, here's the kicker, for them to be able to do this, they need to take a short pause from ramping up their current technology's speed, and moving Dothan over to a bigger mass production line. This is why we find Intel pretty silent right now, and most likely the same with AMD (anticipation; AMD's a very reactionary company).
So, I'm very sure that this is one of the top priorities sitting on the desks of Intel Engineers, I applaud them for taking every step towards a cleaner environment while making my newest tech gadgets.
well, maybe he's just looking for a better tool, maybe ethereal, tcpdump, and etherpeak don't offer a specific feature he's looking for (although, i have no clue what it might be, looking at UDP packet?)
The registry is actually quite a good idea, with a horrible implementation. Basically, it shouldn't be a single file database-style solution. This adds confusion that isn't needed. The other reason it was poorly implemented was because they continued to use INF files, instead of depreciating the entire subsystem, for removal in windows 98 (95 was the first instance of the "system registry" as we've came to know it).
That's probably the whole reason Windows is in as bad of shape as it is; lack of standards, or downright bastardization, reinvention, or some combination of the two.
In theory, they could fix the registry, and I'm working on a similar Linux project right now. The idea of having directories redefineable (think: XML file; sorta what microsoft did with Desktop.ini) based on content, and virtual directories that have any content that fits a certain layout (think: Virtual MP3 Directory, Virtual Movie Directory, Virtual Config file Directory). Defining a virtual conf dir would allow you a one stop location to fix whatever system problem you wanted, and you could even make a purdy little gtk2 or qt3 app made specifically for editing these files (although, it would end up becoming something like G-Edit with a Sidebar).
Of course, this can be implemented by microsoft too, since their new WinFS will be an SQL-based file system. Using a Redefineable directory like "C:\Windows\Configuration Files\" could eventually replace the "registry", and help Windows become a whole lot more secure (by setting appropriate permissions on this directory).
Usually, ideas are very sound and awesome pieces of work. It's the implementations that are the bitch...
Debian needs a subdistro with less archs
*COUGH* USERLINUX *COUGH*
uhh... UserLinux
j00 ph4il 17.. it's OpenBSD, not Linux
He'll let you know when he's done....
Sounds like Gay Porn anyway... wait.. KNIGHT RIDER?? I'm GLAD they cancelled that...
*to get the joke, note my name.*
The minute Microsoft pushes their crap into a bios chip is the minute I move to Apple computers. Mac's are better anyways [/flameoff]
LinuxBIOS really isn't for everyone. It's not as hardware agnostic as a typical harddrive based bootloader is, simply because a bios chip has a tiny amount of eeprom (256k). While this is suffiecient for an optimized system (cluster node, web server, invariant hardware, etc), it's really not good for a home user who'd love to throw his/her new radeon in and not have to insmod their driver everytime they bootup. It's also hell to recompile the kernel, as one false move and your bios is history.
One thing that may be beautiful: flash ram being used instead of current bios archetecture. Flash ram is cheap, and spacious, and can hold quite a large kernel-image. But this gets into more of what XIP is trying to do....
I understand your concerns of Linux becoming Microsoft-ish, and locking in people, but come on, it's just a BIOS image. Don't want it, go to the vendor's website, download the newest one, write it on. Of course this is more dangerous, but for that same reason, I don't believe we'll be seeing a huge push for LinuxBIOS any time soon.
Errr... I really need to re-read what I write. It's a different THEORY, but similar execution.. eXecution In Place is speed via giving the computer EXACTLY what it needs to run, whereas LinuxBIOS is simply skiping the step of harddrive start. Completely different in the fact they're starting from opposite ends of the same problem, but slight in the fact that they've reached an almost common ground.
Fuck grub, this is moving the whole kernel into flash-ram. This is to save the step of having to connect to the hard drive until when it's actually needed to run software, which can be done WHILE the system is booting the kernel, instead of before-hand. And as we know, in the computer world, multitasking is a good way of speeding up things.
;)
Well, for starters if my understanding of eXecute In Place is holding, it's completely different. The ROM chip holds a compressed kernel, that's decompressed on the fly and used to start the system. The only thing this skips is the need to do a startup seek for harddrives, and that whole step of initalizing slow hardware (ie, Hard Disk platters). eXecute In Place is intented to simply run the program natively, ie uncompressed kernel image sitting in ram, so it can just be thrown straight into the processor and executed. The difference is slight, but none-the-less important.
Not from how I understand it. This is actually saving a Kernel-Image onto the bios chip itself, therefore booting straight to the kernel from the chip, basically skipping the step of "BIOS tells HD to find MBR, then read the data from MBR to ram, execute MBR instruction". Or at least, that's how I understand it?
lol, love the site!
;)
But once again, I can reaffirm to you that I'm choosing my profession because I love it.. (hate perl, but that's a whole 'nother story... damned syntatically obsfusecated language...real men use python
Sure, as you all can call me young, you can cite my inexperience, but you can't cite my intelegence, nor can you cite my ambishions nor my abilities. In 4 years, 6 years, 10 years from now, I'll be happy one way or another, because in one way or another, there will always be a niche that needs to be filled where outsourcing is just not an option, even if this does mean I'm not doing super cool innovative work, and not something more than a site manager or something like that. The fact is, there will always be jobs in computers, finding them may be hard, but if you love your work as much as I do, than you have the motivation required.
;), hate credit cards (too much power corrupts), oh, and did I mention I loved my work with computers?
Oh, and about the mortgage, 2.5 kids, an ex-wife, SUV thing: I'm a firm believer in using condoms, won't get married for a long time *trust me, I've had my share of bad relationships at this age to know that no Sure Thing (tm) is worth the impulse.., an enviromentalist (as much as I can be...), will drive a Gas Electric car as soon as I can afford to purchase one
Sorry your life crapped out dude, but I refuse to run mine like that. And I'm worried too about the shift, it's not that I'm not worried, it's that times are changing. I'm going to do computers because I love computers, not because of money, not because of anything else. I love the ability to extend the mind into a chip, to do work at a rate unfathomable by most humans, and the ability to improve our lives that exists within them. These are my aspirations, my goals while working with computers, getting rich is about Null on that list.
One last note: I didn't say live with it, I said "adapt". Change, become something new, be dynamic, force change. Innovate, make your superiors notice you, do work that truly becons being done, not work that's painful and agonizing. Don't mourn your life, LIVE IT!!!!!
I'm to be a freshman in the fall as well, and I'm still going to Major in Computer Engineering, but only because I want to do something that I truely love, and really don't care about the pay. The truth is these jobs are going over seas, and they're moving quickly, but as we've seen with most of these job fluctuations, they tend to be short term (think: NAFTA and the like...), and they tend to stablize themselves quickly. Worst comes to worse, I'll move to India ;)
I think the biggest reason today that jobs are shifting overseas is simply the costs of running a redundant business. Very few companies are actually innovating these days, and those that are, do their work in the good ol' USA because of strong patenting laws (yes, too strong, we know..) and the like. Those same companies are offsetting the price of innovation by reducing the cost of tech support, sending it offshores to cheaper labor. I think the best way to get out of this is simply a change in buisness model; too many buisnesses are worried about the upfront costs as compared to the long-range profits to be gained, and are getting downright greedy and stingy when it comes to money...
Basically, the economic structure of America is changing. Don't like it? Move. Or stay here and adapt.
Whoa... wait... does this mean that more than one programmer actually wrote Windows XP?
I'm SHOCKED!!!!