Like atheletes, engineers are born. If you picked the field for the big money and not getting your hands dirty, you will never be able to compete against those of us who were born to it.
Chip design, like compiler design, is one of those things that has a reputation for being much more difficult that it really is, primarily because even among technophiles there's little understanding of what's involved.
Realize that undergraduate students routinely design and simulate RISC CPUs as part of a semester course. Sure, we're not talking state of the art here, but the principles are the same. Get a group of professionals together, and designing a more modern CPU is doable.
Also note that much of the complexity in chips like the x86 comes from:
* having to support 300+ instructions, a large chunk of which are rarely used * along the same lines, complex oddities like MMX * support for 16-bit mode * trying to make a 22+ year old CISC processor with 8 registers be really fast * more bits of historical wackery, like an 80-bit FPU
and so on. Starting clean, especially if you aren't trying to push the envelope right away, makes things much more approachable.
2) By copying someone else's design, youre relegating your work to a "second place" not-quite-as-good-as-the-original monicker.
Yup. This is the same mistake that "underdogs" always make. The latter years of the Amiga were so depressing, with everyone trying to clone MS-DOS and Genesis/SNES games. Ditto for the latter years of the Apple IIgs. The authors of such games thought they were showing the world that these lesser known machines were as good as the popular ones, but it just made them look second rate.
You get someone to switch by doing something exciting and fresh.
The typical game has a budget of 4-8 million US dollars, is designed by committee, and is created by 30-100 people working overtime for years. I know that fanboys think that there's a Designer With A Vision behind each game, and so on, but the reality is that marketing concerns override all else (games are marketed as toys) and there's no equivalent of the art film (meaning a movie created without bending to marketing concerns, even though it greatly lowers the odds of being successful).
It has always been crap, and even it's creator has stated that it's not a good game. Just because something is rare doesn't mean it's worth collecting.
Miyamoto's games--and Nintendo's games in general--are targeted toward all ages, with the exception of a group from about 15-25 who reject them as being remnants of the childhood they have left behind. The typical pattern is:
Ages 6-14: "These games are great fun!" Ages 15-25: "Nintendo games are for kids! I want mature games!" [Where "mature" means either dark science fiction or extreme violence.] Ages 26+: "These games are great fun!"
It's been amusing to watch how often this occurs. You can see kids hit an age where the reject Mario and flock toward stuff like Resident Evil. Then after a while they work through it all and just play what's fun. Most of the time. Some people never give it up:)
I love projects like this. May just the fact that they succeeded inspire others to try the same thing.
On a tangent, this is the kind of thing that's been very possible for some time, but most people blindly assume that it's much too difficult. Similarly, writing a compiler for a high level language is a relatively easy project. You could do it in a semester course, or a month of spare time, but mention "writing a compiler" to the great majority of programmers, even those with lots of experience, and they run away in terror.
Personal chip design reached critical mass back around 1994.
Wow, the misinformation and narrow-mindedness has been flying on this topic!
I see a lot of stock "What, again!?" responses. Realize that Moore's law could continue, but we're past the point of diminishing returns. It's gotten to where Intel releases a chip that has a clock increase of 9%, resulting in a benchmark increase of 4%, for a 12% increase in power consumption (and a price increase 100%). This obviously is not a good road to continue down.
But realize, and this may come as a shock if you think that Processor = Intel and AMD and maybe PowerPC, it's not that difficult for a college student to design, implement, and test an FPGA processor that outruns a Pentium 4 for certain, specialized tasks. It's also generally accepted that 90% or more of the transistor space in a "modern" commercial CPU is not actively being used for computation: instruction cache, data cache, branch prediction cache, and so on. It gets worse if you consider all the space for x86 instructions that rarely get used, like all the old segmented mode stuff, MMX, and the huge number of transistors required to implement an 80-bit floating point unit. Floating point is good, yes, but even the 3D games which are what really pushed the need for floating point are just fine with 32-bit single precision floats. Not to mention that everyone else stops at 64-bits.
The point I'm driving towards is that we could do a lot more with current processor design technology if we weren't expanding in all directions at once, trying to design a general purpose processor that runs at 4GHz. You can greatly release the pressure by looking at exactly what kinds of operations modern software needs and designing a chip around that. There have been a number of real cases where a simple, FPGA processor beats off-the-shelf hardware by factors of 30 or more.
For example, C and C++ have been steadily on the decline for a number of years now. Why write C when you can write Python? You can even write some impressive games entirely in Python these days. Or why not move on to more reasearchy languages that used to be laughably slow in the MS-DOS days, but but fly on any computer released since 1998? The oft-cited reason C is still used is because it's a system-level language that matches the hardware. But what if the hardware matched something else?
I'll be glad to see the end of the current desktop PC fiasco, even though it may result in turmoil for a while.
AI is a euphemism for "behavior." When I hear people complaining about how games aren't using the latest in AI research, I want to respond "that's because games don't really use AI" at least not what people think of as AI. AI in a typical game is just a list of weighted rules, such as "if the player has a more powerful weapon than character X, make character X run away." When you have lots of such rules and you twiddle with them a lot, then you get so-called AI.
Putting in random factors makes things much harder to pin down. Maybe when a character spots you, there will be a 50% "run or attack" decision. If the decision to run, then you think "Ha, ha, ha, he's running scared!" If the decision is to attack, and he gets you, then you think "Wow, that guy was good." If he attacks and you get him, then you feel like you're doing well.
To a great extent AI is psychological. You read into things what you want.
possibly the only person with both a Ph.D. in computer science and an MFA in poetry)
That's wonderful! For some time now I've been thinking that perhaps a computer science degree is exactly what geeks don't need. Heck, they're already wrapped up in the tech world, and they'll spend the rest of their lives coding, so while not get well rounded early on. Get a degree in history or literature or creative writing, then get the computer science degree later.
The uber-geeks are often the most stubborn, the most prone to get into Slashdot arguments, the ones who have the narrowest views. The more interesting techies with wider views often tend to get out of technical fields later on, because mindless code monkeys who think C++ is The Way and develop software by working 12-14 hour days, well, they're just so mind numbing after a while.
This means there's a lot of potential to teach them about Linux, *BSD
First off, I'm being sincere here. I am not trolling.
Now really, do we want to be introducing kids to what will, in all liklihood, be yesterday's technology in ten or twenty years? A lot of people in computer science are pretty shocked that UNIX-alikes ended up being so hot in the 21st century. Twelve years ago, UNIX was rapidly on its way out and "good riddance" was the general consensus. Even if you disagree with that, then look as it another way: You want kids to get excited about something bright and fresh that lets them move ahead. UNIX is not what you want to focus on here.
Perhaps the most unfortunate part of Linux has been that it allowed 1970s UNIX-heads to get a second chance to live their glory days. It's as if all of a sudden the Commodore 64 became the basis for millions of cell phones, and the C64 hackers of yore could all start coding 6502 assembly again. That's good for them, but it's not an advance. And, sadly, so many people jumping on the Linux bandwagon seem to want to argue themselves to death over the status quo being the ultimate pinnacle of computer. Heck, you can't even get rid of X Windows, because there's always a group who has to bring up it's network transparency (something that less than 1% of all Linux users make regular use of). We're running in circles, scared of anything else that might come along. There needs to be some real excitement for these kids, not just the knowledge that they can be 'leet by hacking configuration files. Teach 'em Smalltalk. Teach 'em REBOL. Teach 'em UnrealScript. Teach 'em anything that lets them get results quickly.
...outsold the motion-picture industry by a billion dollars last year, and movie studios and record labels wonder why they are losing money?
That games made more money than movies is a popular myth. Games make more money than the box office, yes, which isn't too surprising considering that games cost $30-50 and movie tickets cost $6-9. But if you factor in VHS and DVD sales, then games are way, way behind.
You know what, I am so sick of people writing boring-ass essays just re-iterating the same things that everyone always says in essays on copyright.
Amen to that. C'mon, being a Student Who Has Been Reading Geek News Web Sites for Gosh Maybe Three Years Now is not any kind of qualification to be a legal mouthpiece.
Re:FUD disguised as a technical comment.
on
AMD's 64-bit Plot
·
· Score: 2
This is incorrect. The Hammer "long mode" uses 32 bits as the default data size. 64 bits are only used for pointers and explicitely overridden 64 bit operands.
Now isn't that *exactly* what I said, that all addresses are now 64-bits? I didn't say anything about integers stored in memory.
This is also incorrect. There are numerous well known techniques used in ALU design that makes precious few operations "O(bits)".
Yes, and those precious few include multiplication and division. As another poster pointer out, you can get these down to O(log N) if you trade space for speed. But if you get a 64-bit divide executing faster than a 32-bit divide, then something is seriously wrong. That must be one really awful 32-bit divide.
In any case, there's no doubling in speed to be had by going from 32 to 64 bits. That some people are quibbling over a few cycles in a divide unit is pretty sad overall.
If you have a 64-bit 2 GHz processor and a 32-bit 2 GHz processor, the 64-bit processor is going to be much faster. This speeds up the whole system, not just the rate at which you make giblets fly.
No. That's a myth. As it stands, Pentiums for many years now have sported 64 bit buses and 64-bit FPUs (well, 80-bit CPUS actually), so we're not talking about bus size and FPU width. We're talking about:
1. All addresses being 64-bits. 2. All internal integer registers being 64-bits.
For #1, realize that this is going to greatly increase the data size of many applications. The larger the data size, the higher the chance of cache misses. In general, this is a loss, not a win.
For #2, realize that some integer operations are O(N) where N is the number of bits involved. 64-bit multiplication and division are slower than the same 32-bit operations. Period.
The gain with 64-bit processors is one of address space and nothing more.
Re:What desktop users want to know..
on
AMD's 64-bit Plot
·
· Score: 5, Insightful
For almost anyone out there, it's the only factor when buying a CPU: speed!
Nope. These days it's price. You can barely, oh so barely, tell the difference between 866MHz and 2.4GHZ, and only then when running certain high-end games or 3D modelling packages. Now go over to Dell's site and price a 2.4GHz system. You can easily get something with 256MB and no monitor for US$800. Now upgrade to a 3.06GHz P4. How much does that does that 27% increase in clockspeed cost you? Just over US$1000. And what does it get you? Remembering that clockspeed does not translate directly to more CPU performance, maybe you're getting a 20% across the board improvement, but _man_ are you paying for it, both in cost and power consumption. And was it worth it, for 27% faster than "more speed than I know what to do with?" Probably not (though I realize that all hardware site weenies will absolutely insist that they can feel the difference when browsing the web on such a machine).
A 400 hz cpu can't send enough info to the gpu to fully utilise the graphic card's processing power. Start reading some tech sites.
On modern GPUs with lots of memory, there's no need to transfer every bit of data to the card. You upload your world, the card renders it. The world lives on the card. So do the textures. With 64MB or more, 8:1 texture compression goes a long way. Remember, the PS2 has 4MB of video memory, period.
Ok. So let me get this straight. Developers should stop making more complex games just because you don't want to have to upgrade your system? I take it you're not getting DooM ]|[. The games that tax a system do it for a reason: they look better than console games!
They look better because:
1. A monitor is clearer and higher resolution than a TV. 2. A GeForce 3 or 4 is practically a second CPU, one that's independent of the rest of your system (with 128MB and 8:1 texture compression, bus speed is a non issue) Plug one into a 400MHz system and you'll get the same graphics.
I think that so-caled "high end" PC gamers don't have any kind of handle on what performance means or where it comes from.
Also note that high-end PC games are selling very poorly. Doom 3 will be the exception, because all the fanboys will buy it, but try listing 12 PC games that require hardware T&L and have sold more than 100,000 copies each. Heck, try listing 5. Or 3. High end PC gaming is all but dead.
Hell, I am still running a PIII 1GHz (and no, I did not inherit this PC from my parents). It is just not feesable for most people to slap down nearly a grand for a new PC, when the one you have will run everything pretty damn good. I am sure if you did a/. poll you would realize that most people do not have the latest systems. So therefore, this article is very relevent.
And if you read my post carefully, I said that I couldn't tell the difference between a 866MHz and 2.4GHz for developing 3D games, and game players don't need nearly as much capability.
This article might matter if you're a student who inherited your family's old PC and want some hope of playing new games on it. Otherwise, it's irrelevant. You can buy a nice 2.4GHz Pentium 4 with a GeForce 4 for under $1000. Actually, you can go below $800 pretty easily right now, if you keep your current monitor. PC performance has been leveling out lately, and the capabilities of what now sells for $800 have barely been scratched.
I'm a game developer. I used to do high-end 3D game development on a P2-450 with 128MB of memory. I shipped commercial products on that system. It was a tad slow, but that was more because of swapping than anything else. Then I moved up to a P3-866 with 256MB. Now that was a nice system: very fast, no swapping, no complaints at all. Again, I shipped commercial products on that system, and meanwhile the gamer kiddies were all moaning about how you can't survive without anything below 1.5GHz or whatever. Heck, Dell advertising claims that a "low end" 1.8 GHz P4 is good for internet browsing and email. Then I moved up to the typical 2.4GHz system that everyone is buying these days. I can't really tell the difference. Compile times are faster, but they were pretty quick anyway.
Bottom line: If a game runs crappily on a system like this, then the developer should be put out of business. There's no excuse. Why mess with this stuff anyway, when you can buy the hassle-free game console of your choice for significantly less than a Radeon?
Sure, biochemists might need the massively paralell processing power to do molecular folding analysis, but regular joe bloes will, IMHO, be very comfortable with quad 2GHz HT Pentium 4s... for a decade at least.
True, but it sure would be nice to:
1. Get the power consumption way, way down--say, to 1W or less.
2. Along the same lines, get rid of the processor fans and heak sinks and pave the way for a much smaller form factor.
Slashdot Mirror
Like atheletes, engineers are born. If you picked the field for the big money and not getting your hands dirty, you will never be able to compete against those of us who were born to it.
Wow, is that pretentious or what?
Chip design, like compiler design, is one of those things that has a reputation for being much more difficult that it really is, primarily because even among technophiles there's little understanding of what's involved.
Realize that undergraduate students routinely design and simulate RISC CPUs as part of a semester course. Sure, we're not talking state of the art here, but the principles are the same. Get a group of professionals together, and designing a more modern CPU is doable.
Also note that much of the complexity in chips like the x86 comes from:
* having to support 300+ instructions, a large chunk of which are rarely used
* along the same lines, complex oddities like MMX
* support for 16-bit mode
* trying to make a 22+ year old CISC processor with 8 registers be really fast
* more bits of historical wackery, like an 80-bit FPU
and so on. Starting clean, especially if you aren't trying to push the envelope right away, makes things much more approachable.
2) By copying someone else's design, youre relegating your work to a "second place" not-quite-as-good-as-the-original monicker.
Yup. This is the same mistake that "underdogs" always make. The latter years of the Amiga were so depressing, with everyone trying to clone MS-DOS and Genesis/SNES games. Ditto for the latter years of the Apple IIgs. The authors of such games thought they were showing the world that these lesser known machines were as good as the popular ones, but it just made them look second rate.
You get someone to switch by doing something exciting and fresh.
Yes there is; Ico.
No, Ico was just like any other mass market, corporate produced game, except that it didn't click with consumers. That's something different.
The typical game has a budget of 4-8 million US dollars, is designed by committee, and is created by 30-100 people working overtime for years. I know that fanboys think that there's a Designer With A Vision behind each game, and so on, but the reality is that marketing concerns override all else (games are marketed as toys) and there's no equivalent of the art film (meaning a movie created without bending to marketing concerns, even though it greatly lowers the odds of being successful).
This reminds me of Miner 2049'er, which seemed to be on almost every platform, but totally sucked.
Funnily enough, I was addicted to that game for a number of months. The Atari 800 version was great. It's generally remembered as a classic.
Chase the Chuckwagon, though, didn't have anything going for it except the dog food tie-in.
It has always been crap, and even it's creator has stated that it's not a good game. Just because something is rare doesn't mean it's worth collecting.
Miyamoto's games--and Nintendo's games in general--are targeted toward all ages, with the exception of a group from about 15-25 who reject them as being remnants of the childhood they have left behind. The typical pattern is:
:)
Ages 6-14: "These games are great fun!"
Ages 15-25: "Nintendo games are for kids! I want mature games!" [Where "mature" means either dark science fiction or extreme violence.]
Ages 26+: "These games are great fun!"
It's been amusing to watch how often this occurs. You can see kids hit an age where the reject Mario and flock toward stuff like Resident Evil. Then after a while they work through it all and just play what's fun. Most of the time. Some people never give it up
I love projects like this. May just the fact that they succeeded inspire others to try the same thing.
On a tangent, this is the kind of thing that's been very possible for some time, but most people blindly assume that it's much too difficult. Similarly, writing a compiler for a high level language is a relatively easy project. You could do it in a semester course, or a month of spare time, but mention "writing a compiler" to the great majority of programmers, even those with lots of experience, and they run away in terror.
Personal chip design reached critical mass back around 1994.
90% of the reviews I see fall into one of these categories:
1. 5 star reviews from crazy fan-types who absolutely love some movie/musician/cartoon, etc.
2. Anti-fans who post 1 star reviews about stuff they can't stand being popular.
3. Fans who've turned into anti-fans, claiming that item X "isn't as good as their previous efforts" and that the creator "has sold out."
Wow, the misinformation and narrow-mindedness has been flying on this topic!
I see a lot of stock "What, again!?" responses. Realize that Moore's law could continue, but we're past the point of diminishing returns. It's gotten to where Intel releases a chip that has a clock increase of 9%, resulting in a benchmark increase of 4%, for a 12% increase in power consumption (and a price increase 100%). This obviously is not a good road to continue down.
But realize, and this may come as a shock if you think that Processor = Intel and AMD and maybe PowerPC, it's not that difficult for a college student to design, implement, and test an FPGA processor that outruns a Pentium 4 for certain, specialized tasks. It's also generally accepted that 90% or more of the transistor space in a "modern" commercial CPU is not actively being used for computation: instruction cache, data cache, branch prediction cache, and so on. It gets worse if you consider all the space for x86 instructions that rarely get used, like all the old segmented mode stuff, MMX, and the huge number of transistors required to implement an 80-bit floating point unit. Floating point is good, yes, but even the 3D games which are what really pushed the need for floating point are just fine with 32-bit single precision floats. Not to mention that everyone else stops at 64-bits.
The point I'm driving towards is that we could do a lot more with current processor design technology if we weren't expanding in all directions at once, trying to design a general purpose processor that runs at 4GHz. You can greatly release the pressure by looking at exactly what kinds of operations modern software needs and designing a chip around that. There have been a number of real cases where a simple, FPGA processor beats off-the-shelf hardware by factors of 30 or more.
For example, C and C++ have been steadily on the decline for a number of years now. Why write C when you can write Python? You can even write some impressive games entirely in Python these days. Or why not move on to more reasearchy languages that used to be laughably slow in the MS-DOS days, but but fly on any computer released since 1998? The oft-cited reason C is still used is because it's a system-level language that matches the hardware. But what if the hardware matched something else?
I'll be glad to see the end of the current desktop PC fiasco, even though it may result in turmoil for a while.
AI is a euphemism for "behavior." When I hear people complaining about how games aren't using the latest in AI research, I want to respond "that's because games don't really use AI" at least not what people think of as AI. AI in a typical game is just a list of weighted rules, such as "if the player has a more powerful weapon than character X, make character X run away." When you have lots of such rules and you twiddle with them a lot, then you get so-called AI.
Putting in random factors makes things much harder to pin down. Maybe when a character spots you, there will be a 50% "run or attack" decision. If the decision to run, then you think "Ha, ha, ha, he's running scared!" If the decision is to attack, and he gets you, then you think "Wow, that guy was good." If he attacks and you get him, then you feel like you're doing well.
To a great extent AI is psychological. You read into things what you want.
possibly the only person with both a Ph.D. in computer science and an MFA in poetry)
That's wonderful! For some time now I've been thinking that perhaps a computer science degree is exactly what geeks don't need. Heck, they're already wrapped up in the tech world, and they'll spend the rest of their lives coding, so while not get well rounded early on. Get a degree in history or literature or creative writing, then get the computer science degree later.
The uber-geeks are often the most stubborn, the most prone to get into Slashdot arguments, the ones who have the narrowest views. The more interesting techies with wider views often tend to get out of technical fields later on, because mindless code monkeys who think C++ is The Way and develop software by working 12-14 hour days, well, they're just so mind numbing after a while.
This means there's a lot of potential to teach them about Linux, *BSD
First off, I'm being sincere here. I am not trolling.
Now really, do we want to be introducing kids to what will, in all liklihood, be yesterday's technology in ten or twenty years? A lot of people in computer science are pretty shocked that UNIX-alikes ended up being so hot in the 21st century. Twelve years ago, UNIX was rapidly on its way out and "good riddance" was the general consensus. Even if you disagree with that, then look as it another way: You want kids to get excited about something bright and fresh that lets them move ahead. UNIX is not what you want to focus on here.
Perhaps the most unfortunate part of Linux has been that it allowed 1970s UNIX-heads to get a second chance to live their glory days. It's as if all of a sudden the Commodore 64 became the basis for millions of cell phones, and the C64 hackers of yore could all start coding 6502 assembly again. That's good for them, but it's not an advance. And, sadly, so many people jumping on the Linux bandwagon seem to want to argue themselves to death over the status quo being the ultimate pinnacle of computer. Heck, you can't even get rid of X Windows, because there's always a group who has to bring up it's network transparency (something that less than 1% of all Linux users make regular use of). We're running in circles, scared of anything else that might come along. There needs to be some real excitement for these kids, not just the knowledge that they can be 'leet by hacking configuration files. Teach 'em Smalltalk. Teach 'em REBOL. Teach 'em UnrealScript. Teach 'em anything that lets them get results quickly.
Basically, Hammer (64 bit cpus) = larger addressing space, not nescessarily faster processing.
And almost certainly higher cost and higher power consumption.
...outsold the motion-picture industry by a billion dollars last year, and movie studios and record labels wonder why they are losing money?
That games made more money than movies is a popular myth. Games make more money than the box office, yes, which isn't too surprising considering that games cost $30-50 and movie tickets cost $6-9. But if you factor in VHS and DVD sales, then games are way, way behind.
You know what, I am so sick of people writing boring-ass essays just re-iterating the same things that everyone always says in essays on copyright.
Amen to that. C'mon, being a Student Who Has Been Reading Geek News Web Sites for Gosh Maybe Three Years Now is not any kind of qualification to be a legal mouthpiece.
This is incorrect. The Hammer "long mode" uses 32 bits as the default data size. 64 bits are only used for pointers and explicitely overridden 64 bit operands.
Now isn't that *exactly* what I said, that all addresses are now 64-bits? I didn't say anything about integers stored in memory.
This is also incorrect. There are numerous well known techniques used in ALU design that makes precious few operations "O(bits)".
Yes, and those precious few include multiplication and division. As another poster pointer out, you can get these down to O(log N) if you trade space for speed. But if you get a 64-bit divide executing faster than a 32-bit divide, then something is seriously wrong. That must be one really awful 32-bit divide.
In any case, there's no doubling in speed to be had by going from 32 to 64 bits. That some people are quibbling over a few cycles in a divide unit is pretty sad overall.
If you have a 64-bit 2 GHz processor and a 32-bit 2 GHz processor, the 64-bit processor is going to be much faster. This speeds up the whole system, not just the rate at which you make giblets fly.
No. That's a myth. As it stands, Pentiums for many years now have sported 64 bit buses and 64-bit FPUs (well, 80-bit CPUS actually), so we're not talking about bus size and FPU width. We're talking about:
1. All addresses being 64-bits.
2. All internal integer registers being 64-bits.
For #1, realize that this is going to greatly increase the data size of many applications. The larger the data size, the higher the chance of cache misses. In general, this is a loss, not a win.
For #2, realize that some integer operations are O(N) where N is the number of bits involved. 64-bit multiplication and division are slower than the same 32-bit operations. Period.
The gain with 64-bit processors is one of address space and nothing more.
For almost anyone out there, it's the only factor when buying a CPU: speed!
Nope. These days it's price. You can barely, oh so barely, tell the difference between 866MHz and 2.4GHZ, and only then when running certain high-end games or 3D modelling packages. Now go over to Dell's site and price a 2.4GHz system. You can easily get something with 256MB and no monitor for US$800. Now upgrade to a 3.06GHz P4. How much does that does that 27% increase in clockspeed cost you? Just over US$1000. And what does it get you? Remembering that clockspeed does not translate directly to more CPU performance, maybe you're getting a 20% across the board improvement, but _man_ are you paying for it, both in cost and power consumption. And was it worth it, for 27% faster than "more speed than I know what to do with?" Probably not (though I realize that all hardware site weenies will absolutely insist that they can feel the difference when browsing the web on such a machine).
A 400 hz cpu can't send enough info to the gpu to fully utilise the graphic card's processing power. Start reading some tech sites.
On modern GPUs with lots of memory, there's no need to transfer every bit of data to the card. You upload your world, the card renders it. The world lives on the card. So do the textures. With 64MB or more, 8:1 texture compression goes a long way. Remember, the PS2 has 4MB of video memory, period.
Ok. So let me get this straight. Developers should stop making more complex games just because you don't want to have to upgrade your system? I take it you're not getting DooM ]|[. The games that tax a system do it for a reason: they look better than console games!
They look better because:
1. A monitor is clearer and higher resolution than a TV.
2. A GeForce 3 or 4 is practically a second CPU, one that's independent of the rest of your system (with 128MB and 8:1 texture compression, bus speed is a non issue) Plug one into a 400MHz system and you'll get the same graphics.
I think that so-caled "high end" PC gamers don't have any kind of handle on what performance means or where it comes from.
Also note that high-end PC games are selling very poorly. Doom 3 will be the exception, because all the fanboys will buy it, but try listing 12 PC games that require hardware T&L and have sold more than 100,000 copies each. Heck, try listing 5. Or 3. High end PC gaming is all but dead.
Hell, I am still running a PIII 1GHz (and no, I did not inherit this PC from my parents). It is just not feesable for most people to slap down nearly a grand for a new PC, when the one you have will run everything pretty damn good. I am sure if you did a /. poll you would realize that most people do not have the latest systems. So therefore, this article is very relevent.
And if you read my post carefully, I said that I couldn't tell the difference between a 866MHz and 2.4GHz for developing 3D games, and game players don't need nearly as much capability.
This article might matter if you're a student who inherited your family's old PC and want some hope of playing new games on it. Otherwise, it's irrelevant. You can buy a nice 2.4GHz Pentium 4 with a GeForce 4 for under $1000. Actually, you can go below $800 pretty easily right now, if you keep your current monitor. PC performance has been leveling out lately, and the capabilities of what now sells for $800 have barely been scratched.
I'm a game developer. I used to do high-end 3D game development on a P2-450 with 128MB of memory. I shipped commercial products on that system. It was a tad slow, but that was more because of swapping than anything else. Then I moved up to a P3-866 with 256MB. Now that was a nice system: very fast, no swapping, no complaints at all. Again, I shipped commercial products on that system, and meanwhile the gamer kiddies were all moaning about how you can't survive without anything below 1.5GHz or whatever. Heck, Dell advertising claims that a "low end" 1.8 GHz P4 is good for internet browsing and email. Then I moved up to the typical 2.4GHz system that everyone is buying these days. I can't really tell the difference. Compile times are faster, but they were pretty quick anyway.
Bottom line: If a game runs crappily on a system like this, then the developer should be put out of business. There's no excuse. Why mess with this stuff anyway, when you can buy the hassle-free game console of your choice for significantly less than a Radeon?
Sure, biochemists might need the massively paralell processing power to do molecular folding analysis, but regular joe bloes will, IMHO, be very comfortable with quad 2GHz HT Pentium 4s... for a decade at least.
True, but it sure would be nice to:
1. Get the power consumption way, way down--say, to 1W or less.
2. Along the same lines, get rid of the processor fans and heak sinks and pave the way for a much smaller form factor.