XP will be seen as a huge problem once driver development focuses on the Vista/Windows 7 driver model. This is why it is almost a bad thing for people to get Windows XP on a new computer, because they WILL need to upgrade at some point.
It isn't just about the instruction set, it is the implementation and getting it to work properly. Even if NVIDIA didn't need to worry about patents and such, there is a LOT of work that goes into designing a processor.
So, NVIDIA decides not to worry about patents and licensing and TRIES to make a CPU. It has to be 100 percent compatible where programs do not crash due to a bug in the CPU, then it needs to compete in terms of performance. Hmmmm, why do people feel Intel has an advantage...isn't it all about performance?
So, again, without the legal issues, what are the chances that NVIDIA can jump into the x86/x86-64 market with a new CPU and get a chip that can even compete with AMD processors? AMD has been in the CPU business for a long time now. Another issue is that unlike the days of the 8086/8088, 80286, 80386, 80486, and Pentium, processors have become more and more complex to the point where trying to get into the business will be a LOT more difficult.
Another thing that people forget about is that AMD introduced extra registers with the Athlon 64, and those extra registers can be used to make certain tasks easier since fewer instructions would be needed.
When you run low on registers, it means you need to do a calculation, save it just to free up a register to do another calculation and around and around. Most people may not deal with the actual machine language, but when a compiler is properly tuned for a chip, it SHOULD make use of extra registers and such. This is why having good compiler support is key for both AMD and Intel. In theory, if AMD really devoted more energy toward compiler optimizations, AMD chips would probably run anywhere from 5 to 15 percent faster than we are seeing these days.
You forget that there is a huge move toward 64 bit right now, so even if the older 80386 instruction set can be emulated, it won't be easy to make something that really competes with the current designs.
Think about it, if AMD is having a hard time being competitive(the Phenom 2 can compete, even if it isn't faster than the Intel chips), how is NVIDIA going to make a competitive chip in the x86 space? Intel is trying to get competitive in the graphics market, but is failing miserably. If Intel with all of that research and development money is having problems with a graphics chip, how well do you think NVIDIA will do?
And of course, VIA had an x86 chip that you never see used except in some really rare cases. NVIDIA may do better than Cyrix, but that isn't saying much.
The whole point of dual-core is that it DOES improve system response times for those who are multi-tasking. There may not be a lot of applications that take advantage of a multi-core environment, but that second core sure does increase how responsive a computer is when you are running multiple applications at once.
Just because an older computer may still be useful does not mean that the world should revolve around computers that are that old. Most of us do not live in a third world country where computer ownership is considered unusual. Expecting NEW operating systems to be in use in places where single-core processors will continue to be the standard means you expect piracy. People who can not afford a dual-core processor won't be able to afford Windows 7.
The problem with that perspective is that Windows Vista as well as Windows 7 are designed for more recently designed computers. This means that 2GB or more of system memory really is needed just to avoid choking the OS.
In much the same way that Intel chips with a small amount of cache memory perform MUCH slower than those with a normal amount of cache, you have to look at an operating system in the same way. No sane person with knowledge of modern operating systems would put Vista on a computer with less than 2GB of RAM. The pre-loading of commonly used applications is a part of the normal function of Vista, and as a result, more RAM is needed for things to run well.
So, new OS...how many people are going to install a new OS on an old Pentium 4 system? People put Linux onto an older machine because it is a known fact that Linux has a smaller memory footprint, so older computers can have their lives extended. But, if you plan to really USE that machine, throw a GUI on there, run a bunch of things in the background plus OpenOffice and such, how much RAM do you want in that Pentium 4?
Many/most corporations avoid running multiple platforms, which is why Windows 2000 stayed in use for a long time after Windows XP was released because the older machines would not be able to run XP well without upgrades.
For individuals, do even those in the tech community drop a new OS on their current "main" computer, or do they test on a test machine to make sure things work properly before adopting the new OS as their main?
So, people with a clue about the dangers will have at least a dual-core processor with 2-4GB of RAM and a 250GB hard drive. For graphics/video, ATI or NVIDIA chip or card with DirectX 9 or 10 support.
I would love to see some comparisons about Windows 7 with 32 bit vs. 64 bit comparisons in terms of performance. Also compare with low end(dual core 2.1GHz AMD chips) vs. high end(quad core 3.0GHz chips from BOTH AMD and Intel), etc.
As far as people holding onto their old computers, for $370 or so at Best Buy, you can get a new tower with a dual-core processor and 3GB of RAM. For that sort of price, people really are dropping their old single-core computers to upgrade. Yes, there are people out there that are too worried about the economy or have financial problems to do that upgrade, but if the hard drive dies, or the OS crashes, the cost to repair the old computer vs. get a cheap new computer tower makes it a logical time to upgrade.
The problem really comes down to application compatibility with ancient legacy applications. DOSbox is needed to run a DOS application full screen, just because Microsoft decided to force the applications to run in a window instead of full screen mode? There are some network applications that do not work under Vista due to the changes to the networking design under Vista.
The UI isn't the issue for SOME people, but it does affect some applications. These are the real reasons some people just can not switch to Vista at this time. The UI is a minor issue that most people can deal with.
Of course, the real reason Vista has failed is that the driver stability for the first five months of 2007 was pathetic(NVIDIA being a significant source of these). If you go five months where the drivers are not stable, it gives a VERY bad impression of the OS. Vista at launch was never really that bad, but due to driver problems, people assumed it was the fault of Vista. Yes, there may have been some issues, but nothing all that bad.
The issue always come back around to compatibility and what software people are running. If a business refuses to upgrade their software in a ten year period, then it should be expected they will run into trouble.
There is a point where moving to a fairly recent platform is a good thing as well. Now, except for certain code, most programs designed around Windows XP run well under Vista. The real issue is compatibility in the business world, and places that run old DOS code that just doesn't run under Vista.
If Windows 7 has compatibility issues, VMware and other companies that make virtualization products will be the real winners. Run Windows 98 in a virtual machine if it is really needed. We have dual and quad core processors in every new machine, so all that is needed now is a good VM. Maybe Microsoft should provide a micro-WinXP that comes with Windows 7 Ultimate, just for this purpose.
This is why I am going to make a full hard drive image before I install any programs. If a game puts SecuROM on there, once I am done playing, I just revert to the image, and no more SecuROM garbage on my machine.
Only games with SOME sort of online component work that way though. If a single-player game is good, you don't have things like leader boards or anything like that, and at most, you end up with a launcher that can't display the message of the day.
If you play single-player games and you NEED access to a leader board, then what does that say about the single-player nature of the game?
A part of the software design process is how to break up the main application into the different components. With multi-threading, the design needs to figure out what can be handled in a different thread, and if having a different thread for that function is worth the code administration needed to tie things all together.
Remember, it is fairly easy to make a different thread and have it do what you want it to do. The difficulty is in how to tie the different threads together to make the application work as expected.
The Phenom 2 is faster clock for clock compared to the Phenom 1.
Phenom 2 when the socket AM3 version is released will be compatible with both socket AM2+ as well as socket AM3 motherboards. Obviously, the older Phenom and this first set of Phenom 2 processors will be DDR2 only parts, so putting them in an environment where DDR3 memory will be used just will not work.
As far as prices go, the current pricing is on the initial batch, and going forward, AMD has at least a bit more room to increase clock speeds. As others have said as well, you can make a VERY cheap AMD based system with an AMD 790GX chipset and one of these machines. Just slap a hard drive in the machine, some cheap DDR 2 memory, and a hard drive, and you are set. Intel, no matter that you may be able to throw a cheap machine together, can't provide a decent quality graphics chip, so you MUST go with an NVIDIA chipset board with integrated graphics on the Intel side if you want to go for "low cost systems". Look at the system prices for AMD vs. Intel at this lower price point, and AMD wins.
If you look at the overall history of computers, you get a good idea about why the big name companies can't really do well in the "gaming PC" area.
Back when the PC was running at 4.77MHz, even an increase of 3.23MHz was huge because the jump was a greater percentage increase. As time has moved forward, it has become harder and harder to make such huge jumps forward simply because of progress. Even a ten percent improvement in performance is considered very good these days as a result of this.
So, in these days where dual or quad core 2.3GHz processors are everywhere, it means that if you want to play games, the CPU power really isn't an issue for most people. The game development companies seem to spend more time focused on eye candy, rather than really taking advantage of what all of this CPU power can provide.
Now, the GPU market for the most part has been following this to an extent. For MOST people, including gamers, a single Radeon 4870 is enough to handle the games we play. Sure, there are some extreme players that want more, but for the most part, a $300-$400 video card is doing everything we might want it to do. The price of video cards has dropped like a rock over the past few years, so the need to pay even $500 for a single video card has dropped.
What the software developers need to do to revive the PC as a gaming platform is to really focus on the strengths of the PC, not just selling a console game on the PC and wondering why it doesn't do well. The PC has the potential of not just better GPU power, but also better CPU power. How many games have taken advantage of these things?
Areas the PC game market could really focus on would be the AI layer, roleplaying options(beyond having 3 verbal options that have only 2 different reactions from NPCs), more endings, different game branches beyond the 2-4 different endings that we have seen in games from back in the days of Wing Commander, and so on.
Game development that really takes advantage of the platform is what the consoles shoot for, but why does it seem there have been so few new AAA game titles for the PC? Why do these developers feel that first person shooters are the only types of games that will sell well when the console market attracts this type of player as well?
HP is focused on making a LOT of computers with the same design and selling them in large quantities. Even their gaming PC division has that same mindset. They could learn from companies like Cyberpowerpc where you have a starting point for a system, and then focus on being able to customize the machine to suit the needs of the customer. This includes different cases, CPUs, motherboards(something HP does not let you do), video cards, etc.
Then again, HP is still selling computers with that crappy NVIDIA 6150LE with integrated graphics, even though there are much newer and better chipsets out there. If the mainstream machines can't keep up with changes and improvements in the industry, people will assume that the gaming division may be as slow to adapt as well.
NVIDIA and AMD also have not shown us good reasons to go with multi-GPU systems. When the drivers require updates to make SLI or Crossfire work with a new game, rather than SLI or Crossfire working for EVERYTHING, it means those who play games that are not very popular see no advantage. Do you see DDO(Dungeons and Dragons Online) getting SLI or Crossfire support in the drivers? What about when Dragon Age comes out, do you think the current drivers will provide SLI or Crossfire support until the game sells at least 500,000 copies?
AP classes are/were intended to let high school students get SOME college education, but to get to that point, the students need to be at or near the college level in that subject. What I was suggesting is that from the very earliest levels, each student advances in each subject individually, with certain prereqs needed for classes, but generally replacing the current "grade/year" system.
For example, the student can be in fifth level math while still only being at a level 3 in English, and a level 4 in history. Yes, some students can currently skip a grade, but until the high school level, you almost never see a different level of progression for each subject.
Algebra is in the mathematics side of things, so students don't end up in algebra without the basics, but those bad at algebra might be advanced to a college level in English or history or some other subject. Knowing how to read and write will be a requirement for just about all other subjects, but those who are slow or limited in their English won't be held back in other subjects in my system.
People are constantly talking about the education system falling behind in the USA compared to other countries. A part of this is an entrenched system that needs a significant overhaul, not just some minor changes with extra funding.
When I hear education, I hear how the system needs more money, as if that is the only problem. I propose that the problem is that the system itself needs a change. The teachers are also not the problem(in most cases), but this idea that they need to teach students in a way that may not be the best, but without the freedom to find something that works. So, why NOT change the system, and transition all the current teachers and students over to it?
The exact issue is that the school system is modeled after the one room schoolhouse. The entire concept of grades K-12 needs to be thrown out, and instead just have each student advance in each subject at his/her own pace. In this way, a student who is good in English but needs math help does not get held back or even looked at strangely.
Every student will have strengths and weaknesses, so it should be the norm to be several "grades" higher in one or two subjects, and possibly one or two grades lower in one or two subjects. When there is no stigma to having difficulty with a given subject, students will no longer have to hide the fact, and their needs can be addressed.
With this sort of system, the school system can finally improve. Throwing more money at a system that is clearly broken will not help, but replacing the system with something that will work and then moving students into that new system WOULD.
It is a sad thing when most people are more willing to replace an old but working computer than they are to replace a clearly broken system. The same applies to Social Security, health care, and everything else. Everyone keeps trying to fix something that is broken beyond repair instead of trying to figure out what to replace these old broken systems with.
Computers do not take the place of thinking and understanding, but they do free the mind to work on higher level tasks. If children do not have access to computers, how can they work in a world where the use of the computer is used for 99 percent of the work out there? Basic skills are important, but at the same time, computers have become a basic skill in many or most work environments.
This has a fairly simple solution if you think about it. There is nothing that says that you can't increase the density of the connection to memory. With a greater density aka smaller size, you could have multiple banks of memory on the same memory module.
The question I was addressing was about the whole idea of memory bandwidth and how to improve the connection to memory. Many people think one channel per memory module, which isn't going to give the greatest amount of bandwidth for the space provided. If the fab process can be improved for processors, the link to memory can also be shrunk. Why limit ourselves to the current "stick" method of having memory go in a slot when a socket supporting 1TB of RAM could be implemented in the same space on the motherboard(with a heat sink/fan if needed)?
Multi-channel memory controller is my response to this. Remember how going to a dual-channel memory controller increased the available bandwidth to memory? Having support for even 32 banks of memory could be implemented if the CPU design and connections are there.
You are thinking along the lines of current computers, not of the applications. People keep quoting the old statement that 640KB should be enough memory for anyone, but then repeat the same mistake they quote. Quantity of memory not only goes up, but the way to talk to that memory also evolves over time.
We used to see the CPU to chipset to memory as the way personal computers would work. Since then, AMD moved to an integrated memory controller on their CPUs, and Intel is finally following the example that AMD set. A dual-channel memory controller used to be the exception, not the rule, but now the idea is very common. In time, a 32 channel memory controller will be the standard even in an average home computer. How those channels are used to talk to memory of course remains to be seen, but you get the idea.
You have failed to notice that AMD is already on top of this and can add more memory channels to their processors as needed for the application. This may increase the number of pins the processor has, but that is to be expected.
You may not have noticed, but there is a difference between AMD Opteron and Phenom processors beyond just the price. The base CPU design may be the same, but AMD and Intel can make special versions of their chips for the supercomputer market and have them work well.
In a worst case, with the support from AMD or Intel, a new CPU with extra pins(and an increased die size) could add as many channels of memory support as required for the application. This is another area where spinning off the fab business might come in handy.
And yes, this might be a bit expensive, but have you seen the price of a supercomputer?
While this is true, the differences between the global economy back then and now are so large, the issue has grown in complexity.
For example, the USA has been losing the manufacturing industry over the past few decades. A healthy economy is one where everything can be handled within the country. When you need to import certain critical goods, you open up the chance for a collapse further down the road.
When the price of oil went through the roof, the entire economy slowed down because it was more expensive to do business, and the transportation of EVERYTHING got so much more expensive, it was a driving source of inflation. If the USA had more sources of oil that were home grown, it could have helped in this regard.
The whole issue with the auto industry is also an example of too little industry causes prices to go up because there are fewer people looking to go into the manufacturing sector, the unions have control over most of that workforce. Honestly, if you have a lot of people without a college education looking to work in manufacturing, that gives manufacturing based companies some leverage against the unions. With a shortage of people lacking that college education, the unions have a lot more power. So, the unions can push for that crazy $80/hour, even though the education level may only call for $30/hour for many.
Hard work is one thing, but without the costs associated with college, those without the college education get more money per dollar from their work. I suspect that most people wish that they could get a steady job working for someone else and make $60+/hour without a college degree.
So, it all ties back in. China having so much manufacturing is growing in financial power, and the USA lacking manufacturing is losing that financial power since this country does not have the exports. If China decided to cut exports to the USA, this country would be in deep trouble.
You seem to have missed that current systems tend to use at LEAST two channels when connecting to memory. Fusion will in no way be competition for systems with a dedicated video card, but if the graphics portion of Fusion is good enough, you will end up without needing a graphics controller on the motherboard. So, you end up with a new tier in the overall levels of systems.
Fusion will basically allow for low-end chipsets without graphics to hit the market that will still provide better graphics performance than what you see from a system with Intel integrated graphics. These systems may very well be cheaper and due to the graphics issues faster under DirectX 10 than you may expect. The CPU may be a bit more expensive for the CPU power, but with the decreased chipset/motherboard cost, the overall system cost will probably be lower.
Current integrated graphics will never compete with a dedicated video card, in much the same way that a GPU core added to the CPU will never compete with a dedicated video card. A GPU core added to the CPU MAY compete well with integrated graphics though.
What I was saying is that a GUI adds CPU/GPU overhead to the system, and for a server, the resources would be better spent on what the server SHOULD be doing, being a server for clients.
AMD has been working on their Fusion project for a while now, where there will be a GPU core on the CPU. I don't know what the current status of that project is at this point, but the AMD Fusion brand has been expanding, so I suspect it may be here sooner rather than later.
When it comes down to it though, Microsoft has already had this in mind when they released Vista. The whole "performance rating" benchmark in Vista is based on a certain "minimum" and "recommended" rating. Having DX 10 in software means that the overall rating of a system can be used to counteract the graphics score being sub-standard. To be honest, I wish Microsoft would implement one of these for all versions of DirectX, and the compatibility tab would allow older games to run properly since changes to DX can cause undesired side effects.
If you have people running DX 10 games on a server, then you either have major staffing problems, or the server is nothing more than a glorified workstation. There is ZERO need for DX 10 graphics on a true server, and really, the need for a GUI should be near zero(unless the server software vendor can't figure out how to code for a true server).
Yes, there are good uses for having a GUI available on a server, but for normal maintenance, a command line SHOULD be all that is needed to reduce the overhead. the GUI places on the system.
One of the things that I find interesting is that Obama is really an African-American. He is not linked in any way to black slavery. His father comes from Africa, not from the USA. This is why he does not come across as "an angry black man".
Previous "black" candidates in this country have always had the upbringing that showed that African-American people have had to fight to get to where they are today, and it has affected their perspectives. Some(not all, maybe not even most) have run as a black candidate as their reason to run for public office, rather than as a good or great candidate who happened to be black. This difference in self image and upbringing is HUGE. Those who always have it in the back of their mind that they have to overcome race to do well will always have a different perspective on things.
I agree that many issues brought up during the election were a side-show to divert attention from the main issues, and these diversionary tactics did NOT start just because of the economy, but were there for months before the economy took center stage.
Slashdot Mirror
XP will be seen as a huge problem once driver development focuses on the Vista/Windows 7 driver model. This is why it is almost a bad thing for people to get Windows XP on a new computer, because they WILL need to upgrade at some point.
It isn't just about the instruction set, it is the implementation and getting it to work properly. Even if NVIDIA didn't need to worry about patents and such, there is a LOT of work that goes into designing a processor.
So, NVIDIA decides not to worry about patents and licensing and TRIES to make a CPU. It has to be 100 percent compatible where programs do not crash due to a bug in the CPU, then it needs to compete in terms of performance. Hmmmm, why do people feel Intel has an advantage...isn't it all about performance?
So, again, without the legal issues, what are the chances that NVIDIA can jump into the x86/x86-64 market with a new CPU and get a chip that can even compete with AMD processors? AMD has been in the CPU business for a long time now. Another issue is that unlike the days of the 8086/8088, 80286, 80386, 80486, and Pentium, processors have become more and more complex to the point where trying to get into the business will be a LOT more difficult.
Another thing that people forget about is that AMD introduced extra registers with the Athlon 64, and those extra registers can be used to make certain tasks easier since fewer instructions would be needed.
When you run low on registers, it means you need to do a calculation, save it just to free up a register to do another calculation and around and around. Most people may not deal with the actual machine language, but when a compiler is properly tuned for a chip, it SHOULD make use of extra registers and such. This is why having good compiler support is key for both AMD and Intel. In theory, if AMD really devoted more energy toward compiler optimizations, AMD chips would probably run anywhere from 5 to 15 percent faster than we are seeing these days.
You forget that there is a huge move toward 64 bit right now, so even if the older 80386 instruction set can be emulated, it won't be easy to make something that really competes with the current designs.
Think about it, if AMD is having a hard time being competitive(the Phenom 2 can compete, even if it isn't faster than the Intel chips), how is NVIDIA going to make a competitive chip in the x86 space? Intel is trying to get competitive in the graphics market, but is failing miserably. If Intel with all of that research and development money is having problems with a graphics chip, how well do you think NVIDIA will do?
And of course, VIA had an x86 chip that you never see used except in some really rare cases. NVIDIA may do better than Cyrix, but that isn't saying much.
The whole point of dual-core is that it DOES improve system response times for those who are multi-tasking. There may not be a lot of applications that take advantage of a multi-core environment, but that second core sure does increase how responsive a computer is when you are running multiple applications at once.
Just because an older computer may still be useful does not mean that the world should revolve around computers that are that old. Most of us do not live in a third world country where computer ownership is considered unusual. Expecting NEW operating systems to be in use in places where single-core processors will continue to be the standard means you expect piracy. People who can not afford a dual-core processor won't be able to afford Windows 7.
The problem with that perspective is that Windows Vista as well as Windows 7 are designed for more recently designed computers. This means that 2GB or more of system memory really is needed just to avoid choking the OS.
In much the same way that Intel chips with a small amount of cache memory perform MUCH slower than those with a normal amount of cache, you have to look at an operating system in the same way. No sane person with knowledge of modern operating systems would put Vista on a computer with less than 2GB of RAM. The pre-loading of commonly used applications is a part of the normal function of Vista, and as a result, more RAM is needed for things to run well.
So, new OS...how many people are going to install a new OS on an old Pentium 4 system? People put Linux onto an older machine because it is a known fact that Linux has a smaller memory footprint, so older computers can have their lives extended. But, if you plan to really USE that machine, throw a GUI on there, run a bunch of things in the background plus OpenOffice and such, how much RAM do you want in that Pentium 4?
Many/most corporations avoid running multiple platforms, which is why Windows 2000 stayed in use for a long time after Windows XP was released because the older machines would not be able to run XP well without upgrades.
For individuals, do even those in the tech community drop a new OS on their current "main" computer, or do they test on a test machine to make sure things work properly before adopting the new OS as their main?
So, people with a clue about the dangers will have at least a dual-core processor with 2-4GB of RAM and a 250GB hard drive. For graphics/video, ATI or NVIDIA chip or card with DirectX 9 or 10 support.
I would love to see some comparisons about Windows 7 with 32 bit vs. 64 bit comparisons in terms of performance. Also compare with low end(dual core 2.1GHz AMD chips) vs. high end(quad core 3.0GHz chips from BOTH AMD and Intel), etc.
As far as people holding onto their old computers, for $370 or so at Best Buy, you can get a new tower with a dual-core processor and 3GB of RAM. For that sort of price, people really are dropping their old single-core computers to upgrade. Yes, there are people out there that are too worried about the economy or have financial problems to do that upgrade, but if the hard drive dies, or the OS crashes, the cost to repair the old computer vs. get a cheap new computer tower makes it a logical time to upgrade.
The problem really comes down to application compatibility with ancient legacy applications. DOSbox is needed to run a DOS application full screen, just because Microsoft decided to force the applications to run in a window instead of full screen mode? There are some network applications that do not work under Vista due to the changes to the networking design under Vista.
The UI isn't the issue for SOME people, but it does affect some applications. These are the real reasons some people just can not switch to Vista at this time. The UI is a minor issue that most people can deal with.
Of course, the real reason Vista has failed is that the driver stability for the first five months of 2007 was pathetic(NVIDIA being a significant source of these). If you go five months where the drivers are not stable, it gives a VERY bad impression of the OS. Vista at launch was never really that bad, but due to driver problems, people assumed it was the fault of Vista. Yes, there may have been some issues, but nothing all that bad.
The issue always come back around to compatibility and what software people are running. If a business refuses to upgrade their software in a ten year period, then it should be expected they will run into trouble.
There is a point where moving to a fairly recent platform is a good thing as well. Now, except for certain code, most programs designed around Windows XP run well under Vista. The real issue is compatibility in the business world, and places that run old DOS code that just doesn't run under Vista.
If Windows 7 has compatibility issues, VMware and other companies that make virtualization products will be the real winners. Run Windows 98 in a virtual machine if it is really needed. We have dual and quad core processors in every new machine, so all that is needed now is a good VM. Maybe Microsoft should provide a micro-WinXP that comes with Windows 7 Ultimate, just for this purpose.
This is why I am going to make a full hard drive image before I install any programs. If a game puts SecuROM on there, once I am done playing, I just revert to the image, and no more SecuROM garbage on my machine.
Only games with SOME sort of online component work that way though. If a single-player game is good, you don't have things like leader boards or anything like that, and at most, you end up with a launcher that can't display the message of the day.
If you play single-player games and you NEED access to a leader board, then what does that say about the single-player nature of the game?
A part of the software design process is how to break up the main application into the different components. With multi-threading, the design needs to figure out what can be handled in a different thread, and if having a different thread for that function is worth the code administration needed to tie things all together.
Remember, it is fairly easy to make a different thread and have it do what you want it to do. The difficulty is in how to tie the different threads together to make the application work as expected.
The Phenom 2 is faster clock for clock compared to the Phenom 1.
Phenom 2 when the socket AM3 version is released will be compatible with both socket AM2+ as well as socket AM3 motherboards. Obviously, the older Phenom and this first set of Phenom 2 processors will be DDR2 only parts, so putting them in an environment where DDR3 memory will be used just will not work.
As far as prices go, the current pricing is on the initial batch, and going forward, AMD has at least a bit more room to increase clock speeds. As others have said as well, you can make a VERY cheap AMD based system with an AMD 790GX chipset and one of these machines. Just slap a hard drive in the machine, some cheap DDR 2 memory, and a hard drive, and you are set. Intel, no matter that you may be able to throw a cheap machine together, can't provide a decent quality graphics chip, so you MUST go with an NVIDIA chipset board with integrated graphics on the Intel side if you want to go for "low cost systems". Look at the system prices for AMD vs. Intel at this lower price point, and AMD wins.
If you look at the overall history of computers, you get a good idea about why the big name companies can't really do well in the "gaming PC" area.
Back when the PC was running at 4.77MHz, even an increase of 3.23MHz was huge because the jump was a greater percentage increase. As time has moved forward, it has become harder and harder to make such huge jumps forward simply because of progress. Even a ten percent improvement in performance is considered very good these days as a result of this.
So, in these days where dual or quad core 2.3GHz processors are everywhere, it means that if you want to play games, the CPU power really isn't an issue for most people. The game development companies seem to spend more time focused on eye candy, rather than really taking advantage of what all of this CPU power can provide.
Now, the GPU market for the most part has been following this to an extent. For MOST people, including gamers, a single Radeon 4870 is enough to handle the games we play. Sure, there are some extreme players that want more, but for the most part, a $300-$400 video card is doing everything we might want it to do. The price of video cards has dropped like a rock over the past few years, so the need to pay even $500 for a single video card has dropped.
What the software developers need to do to revive the PC as a gaming platform is to really focus on the strengths of the PC, not just selling a console game on the PC and wondering why it doesn't do well. The PC has the potential of not just better GPU power, but also better CPU power. How many games have taken advantage of these things?
Areas the PC game market could really focus on would be the AI layer, roleplaying options(beyond having 3 verbal options that have only 2 different reactions from NPCs), more endings, different game branches beyond the 2-4 different endings that we have seen in games from back in the days of Wing Commander, and so on.
Game development that really takes advantage of the platform is what the consoles shoot for, but why does it seem there have been so few new AAA game titles for the PC? Why do these developers feel that first person shooters are the only types of games that will sell well when the console market attracts this type of player as well?
HP is focused on making a LOT of computers with the same design and selling them in large quantities. Even their gaming PC division has that same mindset. They could learn from companies like Cyberpowerpc where you have a starting point for a system, and then focus on being able to customize the machine to suit the needs of the customer. This includes different cases, CPUs, motherboards(something HP does not let you do), video cards, etc.
Then again, HP is still selling computers with that crappy NVIDIA 6150LE with integrated graphics, even though there are much newer and better chipsets out there. If the mainstream machines can't keep up with changes and improvements in the industry, people will assume that the gaming division may be as slow to adapt as well.
NVIDIA and AMD also have not shown us good reasons to go with multi-GPU systems. When the drivers require updates to make SLI or Crossfire work with a new game, rather than SLI or Crossfire working for EVERYTHING, it means those who play games that are not very popular see no advantage. Do you see DDO(Dungeons and Dragons Online) getting SLI or Crossfire support in the drivers? What about when Dragon Age comes out, do you think the current drivers will provide SLI or Crossfire support until the game sells at least 500,000 copies?
AP classes are/were intended to let high school students get SOME college education, but to get to that point, the students need to be at or near the college level in that subject. What I was suggesting is that from the very earliest levels, each student advances in each subject individually, with certain prereqs needed for classes, but generally replacing the current "grade/year" system.
For example, the student can be in fifth level math while still only being at a level 3 in English, and a level 4 in history. Yes, some students can currently skip a grade, but until the high school level, you almost never see a different level of progression for each subject.
Algebra is in the mathematics side of things, so students don't end up in algebra without the basics, but those bad at algebra might be advanced to a college level in English or history or some other subject. Knowing how to read and write will be a requirement for just about all other subjects, but those who are slow or limited in their English won't be held back in other subjects in my system.
People are constantly talking about the education system falling behind in the USA compared to other countries. A part of this is an entrenched system that needs a significant overhaul, not just some minor changes with extra funding.
When I hear education, I hear how the system needs more money, as if that is the only problem. I propose that the problem is that the system itself needs a change. The teachers are also not the problem(in most cases), but this idea that they need to teach students in a way that may not be the best, but without the freedom to find something that works. So, why NOT change the system, and transition all the current teachers and students over to it?
The exact issue is that the school system is modeled after the one room schoolhouse. The entire concept of grades K-12 needs to be thrown out, and instead just have each student advance in each subject at his/her own pace. In this way, a student who is good in English but needs math help does not get held back or even looked at strangely.
Every student will have strengths and weaknesses, so it should be the norm to be several "grades" higher in one or two subjects, and possibly one or two grades lower in one or two subjects. When there is no stigma to having difficulty with a given subject, students will no longer have to hide the fact, and their needs can be addressed.
With this sort of system, the school system can finally improve. Throwing more money at a system that is clearly broken will not help, but replacing the system with something that will work and then moving students into that new system WOULD.
It is a sad thing when most people are more willing to replace an old but working computer than they are to replace a clearly broken system. The same applies to Social Security, health care, and everything else. Everyone keeps trying to fix something that is broken beyond repair instead of trying to figure out what to replace these old broken systems with.
Computers do not take the place of thinking and understanding, but they do free the mind to work on higher level tasks. If children do not have access to computers, how can they work in a world where the use of the computer is used for 99 percent of the work out there? Basic skills are important, but at the same time, computers have become a basic skill in many or most work environments.
This has a fairly simple solution if you think about it. There is nothing that says that you can't increase the density of the connection to memory. With a greater density aka smaller size, you could have multiple banks of memory on the same memory module.
The question I was addressing was about the whole idea of memory bandwidth and how to improve the connection to memory. Many people think one channel per memory module, which isn't going to give the greatest amount of bandwidth for the space provided. If the fab process can be improved for processors, the link to memory can also be shrunk. Why limit ourselves to the current "stick" method of having memory go in a slot when a socket supporting 1TB of RAM could be implemented in the same space on the motherboard(with a heat sink/fan if needed)?
Multi-channel memory controller is my response to this. Remember how going to a dual-channel memory controller increased the available bandwidth to memory? Having support for even 32 banks of memory could be implemented if the CPU design and connections are there.
You are thinking along the lines of current computers, not of the applications. People keep quoting the old statement that 640KB should be enough memory for anyone, but then repeat the same mistake they quote. Quantity of memory not only goes up, but the way to talk to that memory also evolves over time.
We used to see the CPU to chipset to memory as the way personal computers would work. Since then, AMD moved to an integrated memory controller on their CPUs, and Intel is finally following the example that AMD set. A dual-channel memory controller used to be the exception, not the rule, but now the idea is very common. In time, a 32 channel memory controller will be the standard even in an average home computer. How those channels are used to talk to memory of course remains to be seen, but you get the idea.
You have failed to notice that AMD is already on top of this and can add more memory channels to their processors as needed for the application. This may increase the number of pins the processor has, but that is to be expected.
You may not have noticed, but there is a difference between AMD Opteron and Phenom processors beyond just the price. The base CPU design may be the same, but AMD and Intel can make special versions of their chips for the supercomputer market and have them work well.
In a worst case, with the support from AMD or Intel, a new CPU with extra pins(and an increased die size) could add as many channels of memory support as required for the application. This is another area where spinning off the fab business might come in handy.
And yes, this might be a bit expensive, but have you seen the price of a supercomputer?
While this is true, the differences between the global economy back then and now are so large, the issue has grown in complexity.
For example, the USA has been losing the manufacturing industry over the past few decades. A healthy economy is one where everything can be handled within the country. When you need to import certain critical goods, you open up the chance for a collapse further down the road.
When the price of oil went through the roof, the entire economy slowed down because it was more expensive to do business, and the transportation of EVERYTHING got so much more expensive, it was a driving source of inflation. If the USA had more sources of oil that were home grown, it could have helped in this regard.
The whole issue with the auto industry is also an example of too little industry causes prices to go up because there are fewer people looking to go into the manufacturing sector, the unions have control over most of that workforce. Honestly, if you have a lot of people without a college education looking to work in manufacturing, that gives manufacturing based companies some leverage against the unions. With a shortage of people lacking that college education, the unions have a lot more power. So, the unions can push for that crazy $80/hour, even though the education level may only call for $30/hour for many.
Hard work is one thing, but without the costs associated with college, those without the college education get more money per dollar from their work. I suspect that most people wish that they could get a steady job working for someone else and make $60+/hour without a college degree.
So, it all ties back in. China having so much manufacturing is growing in financial power, and the USA lacking manufacturing is losing that financial power since this country does not have the exports. If China decided to cut exports to the USA, this country would be in deep trouble.
You seem to have missed that current systems tend to use at LEAST two channels when connecting to memory. Fusion will in no way be competition for systems with a dedicated video card, but if the graphics portion of Fusion is good enough, you will end up without needing a graphics controller on the motherboard. So, you end up with a new tier in the overall levels of systems.
Fusion will basically allow for low-end chipsets without graphics to hit the market that will still provide better graphics performance than what you see from a system with Intel integrated graphics. These systems may very well be cheaper and due to the graphics issues faster under DirectX 10 than you may expect. The CPU may be a bit more expensive for the CPU power, but with the decreased chipset/motherboard cost, the overall system cost will probably be lower.
Current integrated graphics will never compete with a dedicated video card, in much the same way that a GPU core added to the CPU will never compete with a dedicated video card. A GPU core added to the CPU MAY compete well with integrated graphics though.
GUI != GPU
What I was saying is that a GUI adds CPU/GPU overhead to the system, and for a server, the resources would be better spent on what the server SHOULD be doing, being a server for clients.
AMD has been working on their Fusion project for a while now, where there will be a GPU core on the CPU. I don't know what the current status of that project is at this point, but the AMD Fusion brand has been expanding, so I suspect it may be here sooner rather than later.
When it comes down to it though, Microsoft has already had this in mind when they released Vista. The whole "performance rating" benchmark in Vista is based on a certain "minimum" and "recommended" rating. Having DX 10 in software means that the overall rating of a system can be used to counteract the graphics score being sub-standard. To be honest, I wish Microsoft would implement one of these for all versions of DirectX, and the compatibility tab would allow older games to run properly since changes to DX can cause undesired side effects.
If you have people running DX 10 games on a server, then you either have major staffing problems, or the server is nothing more than a glorified workstation. There is ZERO need for DX 10 graphics on a true server, and really, the need for a GUI should be near zero(unless the server software vendor can't figure out how to code for a true server).
Yes, there are good uses for having a GUI available on a server, but for normal maintenance, a command line SHOULD be all that is needed to reduce the overhead. the GUI places on the system.
One of the things that I find interesting is that Obama is really an African-American. He is not linked in any way to black slavery. His father comes from Africa, not from the USA. This is why he does not come across as "an angry black man".
Previous "black" candidates in this country have always had the upbringing that showed that African-American people have had to fight to get to where they are today, and it has affected their perspectives. Some(not all, maybe not even most) have run as a black candidate as their reason to run for public office, rather than as a good or great candidate who happened to be black. This difference in self image and upbringing is HUGE. Those who always have it in the back of their mind that they have to overcome race to do well will always have a different perspective on things.
I agree that many issues brought up during the election were a side-show to divert attention from the main issues, and these diversionary tactics did NOT start just because of the economy, but were there for months before the economy took center stage.