That TV-out issue tends to stem from the fact that some card manufacturers (NOT Nvidia) use Philips TV encoder chips on their boards. Those chips suck (I still can't make the one on my GeForce 4 card work, but that's the card manufacturer's fault, not Nvidia's.) What do you mean by "bashing" the Kyro chipsets? It's natural for a company to extol the virtues of their own products while downplaying those of their competitors - EVERY company that has competitors will do that. At the moment, yes, they are the leading company in the video chipset market. This is more the fault of lackluster development efforts on the part of their competitors - they haven't had a whole lot of competition in the high-end graphics card market. That will probably change with the advent of the Radeon 9700.
Now what's all this about getting sued by 3dfx? I don't remember that...
That would explain why everything runs faster with OpenGL. In general, I don't like the idea of adding extensions to APIs that only support a small subset of available hardware, but if it can be done transparently and in such a way as to not significantly degrade the application when using other hardware, it's fine. It's especially fine when I can get WinAmp visualization plugins to run at 650 FPS...:P
It's entirely possible that the TNT2 has better OpenGL support than the GeForce cards. For what it's worth, GTA3 only ran at about 10 FPS (average) on my Athlon XP 1700+ with 256MB PC2700 and an overclocked GeForce 2 GTS at 640x480x16, low draw distance, all the fancy visual and audio effects off- it's REALLY dependent on DirectX 8 support, which the NV17 core lacks. Try upgrading to a GeForce 3 or 4, or one of the newer Radeons; your framerate will probably quadruple. I did, and it screams at 1280x1024x32 with all the effects turned up.
Okay, that's damned cool. I have to show that to my brother - he has a really, really shitty GeForce2 MX200 on his machine and can barely play GTA3:P. I have noticed that DirectX is quite a bit slower than OpenGL. Even in older games like Half-Life, my framerate drops precipitously if I switch it to DirectX mode. I wish more games would take advantage of OpenGL, as it would make it far easier to port to other platforms by not requiring the developers to implement a different proprietary graphics API. Too bad the Redmond juggernaut is so pushy.
You need good hardware to run it. It's a good port, but they didn't enhance it a lot for the PC. I don't think it'll run well with anything less than a GeForce 4 Ti. And frankly, it's so much easier to play on the PC (better controls, etc.) that it's worth spending $150 for a new video card in order to be able to play it.
That's actually completely untrue. Yellow #5 does not retard genital growth, nor does it act as a spermicide or anything else. I still can't believe how persistent that rumor is...
Yes, they are an opportunity to create better software. Unfortunately, the reviewer's attitude is fairly silly. They claim that the new P4 is the best thing since sliced bread for everything but older legacy applications. However, when they say "older legacy applications", they mean "anything that doesn't use SSE2". It simply is not a fair comparison to say that a CPU is better because the benchmarks optimized for it worked better on it than on a different manufacturer's product. Perhaps the optimized applications should be singled out and labeled as incompatible rather than the older, more standardized ones.
If your processor is running above 60C, it's because *you* have not configured the system properly. My Athlon XP 1700+, overclocked safely and carefully to the speed of a 1900+, rarely reaches temperatures in excess of 40C. I am not doing anything extreme like water cooling. I have a few case fans, and a decent heatsink. Intel systems run hot too if the case isn't properly cooled. I've seen it happen. The fact is that AMD and Intel chips don't run at temperatures that are that different; Intel just designed better ways of dealing with the excess heat.
Keep in mind that RDRAM accesses the chips in serial instead of in parallel, which means that the performance gain from that higher speed is not particularly exceptional. Also, you have to consider that each CPU has different strengths and weaknesses - the Athlon is better at integer math, for example, while the Pentium 4 shines at floating-point.
If you're careful and use a good heatsink/fan, the risk of burning out your CPU is fairly small, even if you do overclock. The reason that Intel chips don't burn out is not because they run cool (they don't), but because their on-board thermal diode automatically clocks the chip down when the temperature gets too high. One of my friends who recently purchased a Pentium 4 chip did not clamp the heatsink down all the way. It did not touch the CPU at all, yet the machine booted and ran at about 400 MHz. I was thoroughly impressed by this, being an AMD guy. AMD's Palomino and Thoroughbred chips have on-board diodes too, but they do not support speed stepping when the CPU overheats. However, *if* you have one of the newer motherboards that can read that diode, it can shut the system down before damage can occur. When people burn out their CPUs, it tends to happen because they screw up when they're installing the heatsink, or because part of the heatsink fails (e.g., the clip holding it to the socket.) AMD's processors certainly run hot, but the risks can be minimized by being careful when working with your system.
Some older Via chipsets (the KT133, for example) had major stability issues. However, their newer releases have shown great improvement over their older chips. One of their spokespersons recently stated that they would not do an A-revision of their KT333 simply because they have been more careful and gotten things right the first time. Via had problems in the past, but their products are now more trustworthy than the chipsets made by SiS, ALi, or any of the other manufacturers. My KT333 system tends to be VERY fast and stable, even under Windows. My KT133A system crashed more. The newer Via chipsets are not dogged by the problems found in their older products.
I forgot to sleep today. Shut up.:P I was attempting to mock the previous post about the benefits when using many, many drives, but I guess the (feeble) humor was lost on you.
ALL audio CD players read discs at 1X. The speed at which the disc was burned does not determine how quickly it can be read - that's entirely dependent on the drive you're reading the disc on, and that speed will be 1X if it's an audio CD player. Audio CDs need to be read at the correct speed in order for the audio to play back correctly and at the right speed. Data CDs can be read at any speed because data files do not always have to be accessed with the same timing.
It's true that SCSI gear is more expensive. However, often the performance gain is quite significant. For average desktop use, there may not be a huge performance gain, but you are quite correct in stating that they are useful for specialized cases. SCSI drives are marvelous for audio/video editing - the seek time is very low and they can move large files around very quickly. I have never seen a buffer underrun occur with a SCSI CD burner. For the average desktop user, SCSI is too expensive to be viable, but it's great if you need high performance.
I was referring to a comment that said those who purchase SCSI drives for workstation use are screwing themselves over. I was not directly referring to the speed of a particular drive.
Each speed increase will turn out more coasters if you're using crappy-ass media that was run over by a truck. If you use good blanks, you won't have any problems. My 24X drive has never burned a coaster, while my 2X model churned out quite a few. I think it has a lot more to do with the quality of the media and equipment than with the speed of the drive.
The data on a stamped CD is stored on an aluminum substrate. It's HIGHLY unlikely that a CD-ROM drive could produce enough heat to cause thermal damage to that substrate. CD-Rs don't use pits at all - they just burn away the dye. They are probably more susceptible to very high temperatures, but I still don't think a drive could reach temps high enough to damage things.
For most uses OTHER than mass replication, this is a fairly insignificant speed increase. A 30-second difference in burn time is not going to make Joe Six-Pack's mix CD any better, nor will it save him a significant amount of time. Yes, it will be a timesaver for production use, but it's not very important for most other people.
Slashdot Mirror
That TV-out issue tends to stem from the fact that some card manufacturers (NOT Nvidia) use Philips TV encoder chips on their boards. Those chips suck (I still can't make the one on my GeForce 4 card work, but that's the card manufacturer's fault, not Nvidia's.) What do you mean by "bashing" the Kyro chipsets? It's natural for a company to extol the virtues of their own products while downplaying those of their competitors - EVERY company that has competitors will do that. At the moment, yes, they are the leading company in the video chipset market. This is more the fault of lackluster development efforts on the part of their competitors - they haven't had a whole lot of competition in the high-end graphics card market. That will probably change with the advent of the Radeon 9700.
Now what's all this about getting sued by 3dfx? I don't remember that...
That would explain why everything runs faster with OpenGL. In general, I don't like the idea of adding extensions to APIs that only support a small subset of available hardware, but if it can be done transparently and in such a way as to not significantly degrade the application when using other hardware, it's fine. It's especially fine when I can get WinAmp visualization plugins to run at 650 FPS... :P
It's entirely possible that the TNT2 has better OpenGL support than the GeForce cards. For what it's worth, GTA3 only ran at about 10 FPS (average) on my Athlon XP 1700+ with 256MB PC2700 and an overclocked GeForce 2 GTS at 640x480x16, low draw distance, all the fancy visual and audio effects off- it's REALLY dependent on DirectX 8 support, which the NV17 core lacks. Try upgrading to a GeForce 3 or 4, or one of the newer Radeons; your framerate will probably quadruple. I did, and it screams at 1280x1024x32 with all the effects turned up.
Okay, that's damned cool. I have to show that to my brother - he has a really, really shitty GeForce2 MX200 on his machine and can barely play GTA3 :P. I have noticed that DirectX is quite a bit slower than OpenGL. Even in older games like Half-Life, my framerate drops precipitously if I switch it to DirectX mode. I wish more games would take advantage of OpenGL, as it would make it far easier to port to other platforms by not requiring the developers to implement a different proprietary graphics API. Too bad the Redmond juggernaut is so pushy.
You need good hardware to run it. It's a good port, but they didn't enhance it a lot for the PC. I don't think it'll run well with anything less than a GeForce 4 Ti. And frankly, it's so much easier to play on the PC (better controls, etc.) that it's worth spending $150 for a new video card in order to be able to play it.
...if it can't run GTA3 *perfectly*, I'll stick with Windows for my silly wastes of time. :P
Code Red is wonderful, but it has an annoying appetite suppressant effect on me. I can never eat much after I chug a bottle of the stuff.
That's actually completely untrue. Yellow #5 does not retard genital growth, nor does it act as a spermicide or anything else. I still can't believe how persistent that rumor is...
Yes, they are an opportunity to create better software. Unfortunately, the reviewer's attitude is fairly silly. They claim that the new P4 is the best thing since sliced bread for everything but older legacy applications. However, when they say "older legacy applications", they mean "anything that doesn't use SSE2". It simply is not a fair comparison to say that a CPU is better because the benchmarks optimized for it worked better on it than on a different manufacturer's product. Perhaps the optimized applications should be singled out and labeled as incompatible rather than the older, more standardized ones.
If your processor is running above 60C, it's because *you* have not configured the system properly. My Athlon XP 1700+, overclocked safely and carefully to the speed of a 1900+, rarely reaches temperatures in excess of 40C. I am not doing anything extreme like water cooling. I have a few case fans, and a decent heatsink. Intel systems run hot too if the case isn't properly cooled. I've seen it happen. The fact is that AMD and Intel chips don't run at temperatures that are that different; Intel just designed better ways of dealing with the excess heat.
Keep in mind that RDRAM accesses the chips in serial instead of in parallel, which means that the performance gain from that higher speed is not particularly exceptional. Also, you have to consider that each CPU has different strengths and weaknesses - the Athlon is better at integer math, for example, while the Pentium 4 shines at floating-point.
If you're careful and use a good heatsink/fan, the risk of burning out your CPU is fairly small, even if you do overclock. The reason that Intel chips don't burn out is not because they run cool (they don't), but because their on-board thermal diode automatically clocks the chip down when the temperature gets too high. One of my friends who recently purchased a Pentium 4 chip did not clamp the heatsink down all the way. It did not touch the CPU at all, yet the machine booted and ran at about 400 MHz. I was thoroughly impressed by this, being an AMD guy. AMD's Palomino and Thoroughbred chips have on-board diodes too, but they do not support speed stepping when the CPU overheats. However, *if* you have one of the newer motherboards that can read that diode, it can shut the system down before damage can occur. When people burn out their CPUs, it tends to happen because they screw up when they're installing the heatsink, or because part of the heatsink fails (e.g., the clip holding it to the socket.) AMD's processors certainly run hot, but the risks can be minimized by being careful when working with your system.
Some older Via chipsets (the KT133, for example) had major stability issues. However, their newer releases have shown great improvement over their older chips. One of their spokespersons recently stated that they would not do an A-revision of their KT333 simply because they have been more careful and gotten things right the first time. Via had problems in the past, but their products are now more trustworthy than the chipsets made by SiS, ALi, or any of the other manufacturers. My KT333 system tends to be VERY fast and stable, even under Windows. My KT133A system crashed more. The newer Via chipsets are not dogged by the problems found in their older products.
I forgot to sleep today. Shut up. :P I was attempting to mock the previous post about the benefits when using many, many drives, but I guess the (feeble) humor was lost on you.
ALL audio CD players read discs at 1X. The speed at which the disc was burned does not determine how quickly it can be read - that's entirely dependent on the drive you're reading the disc on, and that speed will be 1X if it's an audio CD player. Audio CDs need to be read at the correct speed in order for the audio to play back correctly and at the right speed. Data CDs can be read at any speed because data files do not always have to be accessed with the same timing.
It's true that SCSI gear is more expensive. However, often the performance gain is quite significant. For average desktop use, there may not be a huge performance gain, but you are quite correct in stating that they are useful for specialized cases. SCSI drives are marvelous for audio/video editing - the seek time is very low and they can move large files around very quickly. I have never seen a buffer underrun occur with a SCSI CD burner. For the average desktop user, SCSI is too expensive to be viable, but it's great if you need high performance.
I was referring to a comment that said those who purchase SCSI drives for workstation use are screwing themselves over. I was not directly referring to the speed of a particular drive.
I've never had problems with audio discs burned on my 24X Mitsumi. Maybe I'm just lucky.
It's only useful to drive a space shuttle to work if you're using 1000 of them. Then you'll save even more time getting to work!
SCSI drives and controllers are always expensive. However, they can provide an ENORMOUS increase in performance compared to ATA drives.
Each speed increase will turn out more coasters if you're using crappy-ass media that was run over by a truck. If you use good blanks, you won't have any problems. My 24X drive has never burned a coaster, while my 2X model churned out quite a few. I think it has a lot more to do with the quality of the media and equipment than with the speed of the drive.
The data on a stamped CD is stored on an aluminum substrate. It's HIGHLY unlikely that a CD-ROM drive could produce enough heat to cause thermal damage to that substrate. CD-Rs don't use pits at all - they just burn away the dye. They are probably more susceptible to very high temperatures, but I still don't think a drive could reach temps high enough to damage things.
For most uses OTHER than mass replication, this is a fairly insignificant speed increase. A 30-second difference in burn time is not going to make Joe Six-Pack's mix CD any better, nor will it save him a significant amount of time. Yes, it will be a timesaver for production use, but it's not very important for most other people.
There are already 56X and faster CD-ROM drives. Those don't tend to shatter discs. I don't think 40X poses any significant dangers to the media used.
Generally, IDE disks don't really need separate drivers... Under Linux, you shouldn't need anything but mkisofs and cdrecord.