You have two situations here. Most of the "RAID" cards are just a PCI-IDE interface, with special drivers. For example, the Promise FastTrack and the HighPoint chips. Software raid, and it sits on top of the block and IDED layers, that, at least in the linux kernel, already talks the > 160 gig addressing. There, you have no problem.
Note, even if your BIOS doesn't support the new addressing, that's not a problem. If the drive is not your boot drive, you don't need to worry at all. If the drive IS your boot drive, you can tell the chipset that your drive is however large you care (I usually say about 100 megs), and enable LBA. Then you make/boot within the first 100 megs, and keep your kernel there. Once the kernel has booted, you can use the complete drive.
You could use one of these new 200 gig drives on those old, ghetto ISA IDE interfaces, you remember them, they shipped with CDROMS, they had a few jumpers, two 74 hundred TTL chips, and a PAL or a GAL? Yeah, well, it's an IDE interface, the same as those PCI ones, just without the DMA and other functionality.
Anyways, next problem, The SuperTrak's, Escalade's, and Adaptec 2400's. They have custom processors and firmware that actually take all the RAID processing off of your chip effectively Hardware Raid, unlike the Win-Raid above (ala winmodem).
Those are probably fixable with a controller BIOS upgrade, pending, of course, vendor support. People have shown that, given a fast enough processor, your peak throughput from those software RAID cards can surpass the hardware raid solutions. I want to see how that falls, when the processor is stressed.
Anyways, poing being, unless you're using hardware raid, you've got no problems, buddy.
Ahmen! I've got an A21p that's has survived the student life and is still gorgeous. Being carted everywhere in a backpack and thumped around with a big stack of books it's amazing that the whole thing works. Honestly, I doubt I will need to replace this laptop for a long time - gorgeous screen, good processor, great keyboard, etc.
There are a number of Intel E7500 based boards from Tyan and Supermicro. However, IMHO, the Rambus is a better option for the P4, which, after all, was designed for Rambus and takes a hit with SDRAM.
Hi! I have a similar problem. Could someone give me helpful tips on how to enjoy the real world? Do you recommend anywhere to eat better than the local quickie mart and pizza delivery place? What can I do when I'm not plugged in? What are these hobby things that people talk about?
Dude! Enjoy the trip. Don't sweat the internet, cellphone's, etc. Just grab a camera, some friends, and explore! You'll find your local AAA office will have more information, maps (quite useful), and tips than most slashdotters.
You might appreciate a review like this one: http://www.digit-life.com/articles/zoltrixpro6/ They actually take the time to try and gauge the quality of the DACs/ADCs with another, reference soundcard. I thought it was much better than "Joe, the Janitor, likes this better than the other one."
I thought it was interesting that they compared the Audigy to the Midiman Delta 66 (Envy24 based), and then mention that the Midiman Delta 2496 (Envy24 based) is a good match to the Audigy. One of the things that I have been looking for in a sound card, is versatile S/PDIF stream generation. Cards like the Audigy and the Live resample all digital output from 44.1khz to 48khz, a process which is done digitally and can introduce artifacts. On the otherhand, the CMI8738 cards, like the Midiman Delta 2448 or the Zoltrix Nightingale Pro 6 both offer both coax and optical digital in and out, and offer 32, 44.1, and 48 khz output.
And, the advantage of digital output - now, I can buy a nice, good rig for doing the digital to analog conversion and amplification. Personally, I bought a Sony STR-DE675 for it (It certainly delivers more power than your average computer speaker set.)
For a while, I used a $18 FM801 based card, with resampled digital out - which still sounded damn good.
My point is this, With digital soundcards, the most important part is the quality of the chipset. The rf and analog design of the soundcard is inconsequential because the digital sound will either work, or it won't. If anyone is interested in digital audio for playing back DVDs and/or Music, but not playing games - I'd sincerely suggest finding a CMI8738 based card.
Oh, and on cables: I've ran digital audio over 50 feet on hand-terminated speaker wire. You could tell when there was too much interference, because the audio would drop out, and the stereo would display 48Khz PCM again, as it resync'ed the signal. The point? The Nature of the Audio transmitted over the digital audio is not affected by the cables. With Analog transmission, the capacitance and inpedence of the cable and connectors could act as a filter, and degrade the audio. However, that does not hold true with digital audio. If you have a truly awful cable, you're going to have spots of audio, if it plays at all. However, that $300 three foot long piece of gold plated, insulated cable is not going to do any better than a $20 video-cable.
woah, if you look at the pictures, and try to make sense of the text...'TrueServer heeft een gigabit lijntje met de AMS-IX en 100Mbit transit uplinks met Level3, Ebone, Telia en AboveNet.' - are they using a Juniper M20 to route 3 100meg uplinks into a gig patch to their switch? Isn't that just a teensy bit excessive?
I built an 8088 based SBC by hand as my senior design project in high school. It had bugs, but I was able to latch the data bus to a few leds, and make them count up from zero.I could have made something useful, if I'd had more time to spend with it.
All you need is the intel docs, really. There are oodles of schematics on line, as well as a number of books on it.
Oh, wait.. not only did bill gates not say that.. no, he also wasn't even close to responsible for that limit.. the IBM PC designers made that decision when they put the framebuffer at A:0000-A:FFFF. (Yeah, I know Monochrome was at B:8000, but yeah, big deal.)
Basically, IBM engineers decided that from 0:0000 to A:0000 was decided upon as a reasonable amount of ram. That's 10 blocks of 65K, hence, the 640K limit. The rest of the space was used for memory mapped I/O.
Further, Bill would have loved to have not had to use EMM386 - that thing is/was a hack. And, Windows 95 onwards completely rewrite any magic that EMM386 does.
mencoder is retarded easy to use, and ridiculously powerful:
mencoder -dvd 1 -o dvdrip.avi
Automatic decss, dvd rip, and encoding with the algorithm of your choice, including three different divx codecs (windows, libavcodec/ffmpeg, and divx4linux.) They have a shitload of other codecs. I can use mplayer to play.wmf's and.asf's, etc. They have multipass encoding and transcoding. It has built in support for video 4 linux, dvds, etc. As well as can output via SDL, xvideo, X11, svgalib, fbdev, etc. It's efficient enough to let me play 8 simultaneous divx streams on my dual pent iii 866 (serverworks he chipset, scsi disk.) GUI frontend to mplayer, etc. Amazing stuff, IMHO.
>> Gigabit ethernet over copper has the advantage of running over your existing cabling (i.e. cat-5 is fine). This avoids having to muck about with fiber, as fiber is a PITA to maintain yourself (getting optically perfect connections for the fiber jacks is picky).
Actually, the siecor unicam series work really, really well. They use a index of refraction matching gel inside the factory polished terminators. All you have to do is cut'n'crimp. They work great. I haven't ever had to do any splicing though - but, given how well the siecor stuff works, I can't see it being a remarkable problem.
The brunt of the installed computers in the world do very mundane things - they type papers, run spreadsheets, browse the internet, and balance checkbooks. Similarly, most of the cars on the road out there are taking kids to school and soccer practice, driving back and forth from work, and getting groceries. Porsche proved itself on the race-track, Honda proved itself by building cheap, reliable cars.
I think the biggest thing in computing would be a route similar to the game consoles - controlled hardware and software environments. Commodity hardware that runs a straightforward operating environment designed specifically to do common computing tasks. 32 bit processors are ideal, have smaller transistor requirements, simpler busses and simpler logic than 64 bit brothers.
I think the computer market is maturing from glitz and glammour to reliability. People are getting to the point where they realize that a 1 gigahertz computer will type letters, send email, and play dvds just as well as a 2 gHz computer.
Lots of people have tried to create a commodity computer - but, there have been few successes. In fact, the only notable ones are the iMac and iBook. Niether are powerhouses, they don't even play games. But, they're great for a distinct majority of the computer using audience. They're cheap and have tightly controlled hardware. Apple is even trying to control software, they're making quality apps to suit common computing needs. iTunes, iPhoto, etc.
I'm a big fan of Apple for their move to OSX, but I appreciate the elegance of their iMac/iBook solution. They're not making Porsche's, they're trying to make Hondas. So far, they're doing a damn good job. Some people need Porsche's and they buy them. If you need 450 horses and 490 foot pounds of torque, you can get it - but, most people don't need it.
Actually, no. The 8088/8086 are really 16 bit processors. They both had a 20 bit address bus, and had 16 bit wide registers. 8088 had an 8 bit data bus; 8086 had a 16 bit data bus. The 80286 added a primitive mmu to get the 286 protected mode and had a 24 bit address bus and 16 bit data bus. The 80386 had a 32 bit address bus, with the same old 16 bit data bus. The 386 also sported 32 bit registers and had a real mmu. The 16 bit registers, ax, bc, etc. were extended to 32 bits eax, ebx. The 16 bit commands still operated on the lower 16 bits of the registers. (register size is specified in the instruction prefix.) The 80486 had the full 32/32 compliment. (That's why you needed simms in pairs on 486's.) The 8086 to the 286 was a stumble, the 386 was definately an improvement.
However, realistically - the x86 line should have stopped around the time of the 386, 486, or perhaps the first pentiums.
The itanium is pretty disgusting. Seriously stupid flaws, for example: there is no integer division - you HAVE to use the FPU for all divisions, including transferring values to and from floating point registers.
I personally hold that Alphas are, architecturally, the king of 64 bit chips. Unfortunately, they don't have the money behind them that x86 does - I think a lot of the industry laments that. As the Pentium IV proves, with enough money you can make anything fast and cheap. Imagine what we could have done with the Alpha.
Re:Well, that's Itanium, and see what it got them.
on
Inside Intel
·
· Score: 1
Intel doesn't care about efficiency, megahertz, or pretty much anything except... megabucks.
Each context-switch also requires a context save, restore, cache-flush etc. The damage isn't too much worse than a context-switch, which happens thousands of times a second, anyways; but, it can impact maximum throughput, because the preempted task will be postponed.
With this patch the kernel becomes preemptible - meaning, other kernel tasks can stop the current one from executing, execute, finish, and allow the stopped tasks to finish.
Net effect - expensive operations can be suspended for user interactiveness. Can this impact performance, Yes. Noticeably? No.
If you're running a big-ass server, it's probably head-less, anyways - and you won't have any large, interactive processes preempting the kernel for smoothness.
If you're running a workstation, this means that X won't bog down as much when you're running those huge simulations, compiles, etc.
If you're on an embedded device, you can use this to try and get real-time responsiveness. (perhaps not ideal, but, in an embedded situation you have enough control that if you need a better real-time guarantee, you have other options (e.g. rtlinux).)
If you're on a modest, consumer PC - X won't suck as much.
All in all, this is a good idea. In theory, you lose some efficiency making several thousand context switches/second, but that's the price you pay for multi-tasking. Yeah, certain kernel operations may take longer, but, you get a better responsiveness, which - for most people, is a good thing. Most interactive individuals are seldomly pegging their processor at 100% utilization for any worthwhile period of time. (Games are an exception.)
If you're doing digital input and output, remember that the soundcard is now acting as a glorified serial port. The MAudio Delta 2448, for example, is a C-Media 8738, and the MAudio Delta 2496 is an Envy24. The EMU10K1, which powers all the versions of the live, resamples all output to 48khz - it does hurt the output. The CMI-8738, on the other hand, can output at 32khz, 44.1khz, and 48khz without a problem. The best thing is, you can purchase some C-Media cards, with decent (not midiman quality, however) drivers and optical + coxial S/PDIF in/out ports for around $40. The Midiman MAudio Delta 2448 is still around $200.
Don't necessarily avoid game cards - but, shop around and evaluate your needs. I'd trust seperate, individual ADC/DAC pairs coupled to a $40 CMI-8738 over a $200 analog soundcard by itself. (note, the ADCs/DACs themselves may cost more than $200 - but, the point is the sound card itself is secondary.)
But, most definately read what other people have to say about different cards.
In my humble quest to see why Realtek 8139C cards were $10ea, invididually, and some people still charge real money for Intel, 3Com, and Adaptec (starfire) boards, I found that some of the more quality boards had bigger buffers, and more features.
But, not meaning to knock cheap chipsets, the C-Media 8738, the heart of most $30 5.1 soundcards, has a better SPDIF implementation than the SB Live (the C-Media can generate different clock signals, rather than resampling all output to 48khz. This is a big deal, when most streams are 44.1khz) The Midiman 2448, a professional quality Coax and Optical SPDIF card - is just a C-Media 8738, with a $200 pricetag. The Midiman 2496, btw, is just a card with an Envy24 chipset.
So, yeah, I don't have the facts to back it up - but, I know that the 100 meg interfaces in the Cisco firewall products are Intel i82559s. (Open up a PIX 515 sometime - how's that for a $7,000 pent 200?)
Slashdot Mirror
There's measuring beta particles of decaying Krypton: Fourmilab Hotbits
... google.
Then, there's LAVALAMP randomness: LavaRND
Oh, and you could connect a radio to a sparcstation, and use broadcast noise at: Random.org
Hell, you could use a webcam pointing at a staticy TV.
Lots of possibilities. Amazing what you can find with
You have two situations here. Most of the "RAID" cards are just a PCI-IDE interface, with special drivers. For example, the Promise FastTrack and the HighPoint chips. Software raid, and it sits on top of the block and IDED layers, that, at least in the linux kernel, already talks the > 160 gig addressing. There, you have no problem.
/boot within the first 100 megs, and keep your kernel there. Once the kernel has booted, you can use the complete drive.
Note, even if your BIOS doesn't support the new addressing, that's not a problem. If the drive is not your boot drive, you don't need to worry at all. If the drive IS your boot drive, you can tell the chipset that your drive is however large you care (I usually say about 100 megs), and enable LBA. Then you make
You could use one of these new 200 gig drives on those old, ghetto ISA IDE interfaces, you remember them, they shipped with CDROMS, they had a few jumpers, two 74 hundred TTL chips, and a PAL or a GAL? Yeah, well, it's an IDE interface, the same as those PCI ones, just without the DMA and other functionality.
Anyways, next problem, The SuperTrak's, Escalade's, and Adaptec 2400's. They have custom processors and firmware that actually take all the RAID processing off of your chip effectively Hardware Raid, unlike the Win-Raid above (ala winmodem).
Those are probably fixable with a controller BIOS upgrade, pending, of course, vendor support. People have shown that, given a fast enough processor, your peak throughput from those software RAID cards can surpass the hardware raid solutions. I want to see how that falls, when the processor is stressed.
Anyways, poing being, unless you're using hardware raid, you've got no problems, buddy.
Ahmen! I've got an A21p that's has survived the student life and is still gorgeous. Being carted everywhere in a backpack and thumped around with a big stack of books it's amazing that the whole thing works. Honestly, I doubt I will need to replace this laptop for a long time - gorgeous screen, good processor, great keyboard, etc.
$75 this year, but they're paying the speakers, so it should have a better set of talks.
>> No, there are no DDR solutions for the Xeon yet.
Actually, yes, yes there are. Of course, they are registered ECC DDR solutions, but - they are DDR.
Behold, The SuperMicro P4DP6
There are a number of Intel E7500 based boards from Tyan and Supermicro. However, IMHO, the Rambus is a better option for the P4, which, after all, was designed for Rambus and takes a hit with SDRAM.
Hi! I have a similar problem. Could someone give me helpful tips on how to enjoy the real world? Do you recommend anywhere to eat better than the local quickie mart and pizza delivery place? What can I do when I'm not plugged in? What are these hobby things that people talk about?
Dude! Enjoy the trip. Don't sweat the internet, cellphone's, etc. Just grab a camera, some friends, and explore! You'll find your local AAA office will have more information, maps (quite useful), and tips than most slashdotters.
You might appreciate a review like this one: http://www.digit-life.com/articles/zoltrixpro6/
They actually take the time to try and gauge the quality of the DACs/ADCs with another, reference soundcard. I thought it was much better than "Joe, the Janitor, likes this better than the other one."
I thought it was interesting that they compared the Audigy to the Midiman Delta 66 (Envy24 based), and then mention that the Midiman Delta 2496 (Envy24 based) is a good match to the Audigy. One of the things that I have been looking for in a sound card, is versatile S/PDIF stream generation. Cards like the Audigy and the Live resample all digital output from 44.1khz to 48khz, a process which is done digitally and can introduce artifacts. On the otherhand, the CMI8738 cards, like the Midiman Delta 2448 or the Zoltrix Nightingale Pro 6 both offer both coax and optical digital in and out, and offer 32, 44.1, and 48 khz output.
And, the advantage of digital output - now, I can buy a nice, good rig for doing the digital to analog conversion and amplification. Personally, I bought a Sony STR-DE675 for it (It certainly delivers more power than your average computer speaker set.)
For a while, I used a $18 FM801 based card, with resampled digital out - which still sounded damn good.
My point is this, With digital soundcards, the most important part is the quality of the chipset. The rf and analog design of the soundcard is inconsequential because the digital sound will either work, or it won't. If anyone is interested in digital audio for playing back DVDs and/or Music, but not playing games - I'd sincerely suggest finding a CMI8738 based card.
Oh, and on cables: I've ran digital audio over 50 feet on hand-terminated speaker wire. You could tell when there was too much interference, because the audio would drop out, and the stereo would display 48Khz PCM again, as it resync'ed the signal. The point? The Nature of the Audio transmitted over the digital audio is not affected by the cables. With Analog transmission, the capacitance and inpedence of the cable and connectors could act as a filter, and degrade the audio. However, that does not hold true with digital audio. If you have a truly awful cable, you're going to have spots of audio, if it plays at all. However, that $300 three foot long piece of gold plated, insulated cable is not going to do any better than a $20 video-cable.
PCMCIA sound cards are quite common. Whats more important, is what type of audio do you want coming out of your laptop? Just regular analog out, or S/PDIF out? Well, Behold: http://www.digigram.com/products/getinfo.htm?prod_ key=11500
woah, if you look at the pictures, and try to make sense of the text...'TrueServer heeft een gigabit lijntje met de AMS-IX en 100Mbit transit uplinks met Level3, Ebone, Telia en AboveNet.' - are they using a Juniper M20 to route 3 100meg uplinks into a gig patch to their switch? Isn't that just a teensy bit excessive?
I built an 8088 based SBC by hand as my senior design project in high school. It had bugs, but I was able to latch the data bus to a few leds, and make them count up from zero.I could have made something useful, if I'd had more time to spend with it.
All you need is the intel docs, really. There are oodles of schematics on line, as well as a number of books on it.
Oh, wait.. not only did bill gates not say that.. no, he also wasn't even close to responsible for that limit.. the IBM PC designers made that decision when they put the framebuffer at A:0000-A:FFFF. (Yeah, I know Monochrome was at B:8000, but yeah, big deal.)
Basically, IBM engineers decided that from 0:0000 to A:0000 was decided upon as a reasonable amount of ram. That's 10 blocks of 65K, hence, the 640K limit. The rest of the space was used for memory mapped I/O.
Further, Bill would have loved to have not had to use EMM386 - that thing is/was a hack. And, Windows 95 onwards completely rewrite any magic that EMM386 does.
mencoder is retarded easy to use, and ridiculously powerful:
.wmf's and .asf's, etc. They have multipass encoding and transcoding. It has built in support for video 4 linux, dvds, etc. As well as can output via SDL, xvideo, X11, svgalib, fbdev, etc. It's efficient enough to let me play 8 simultaneous divx streams on my dual pent iii 866 (serverworks he chipset, scsi disk.) GUI frontend to mplayer, etc. Amazing stuff, IMHO.
mencoder -dvd 1 -o dvdrip.avi
Automatic decss, dvd rip, and encoding with the algorithm of your choice, including three different divx codecs (windows, libavcodec/ffmpeg, and divx4linux.) They have a shitload of other codecs. I can use mplayer to play
>> Gigabit ethernet over copper has the advantage of running over your existing cabling (i.e. cat-5 is fine). This avoids having to muck about with fiber, as fiber is a PITA to maintain yourself (getting optically perfect connections for the fiber jacks is picky).
Actually, the siecor unicam series work really, really well. They use a index of refraction matching gel inside the factory polished terminators. All you have to do is cut'n'crimp. They work great. I haven't ever had to do any splicing though - but, given how well the siecor stuff works, I can't see it being a remarkable problem.
mov ax, $4c00
int 21
Doesn't this ever get old? Do we really care how to return to the command interpreter?
Shoot The Other Machine In The Head. An Explanation. I'm serious.
Sorry, the DX had a 32 bit data bus, the SX had a 16 bit data bus.
The brunt of the installed computers in the world do very mundane things - they type papers, run spreadsheets, browse the internet, and balance checkbooks. Similarly, most of the cars on the road out there are taking kids to school and soccer practice, driving back and forth from work, and getting groceries. Porsche proved itself on the race-track, Honda proved itself by building cheap, reliable cars.
I think the biggest thing in computing would be a route similar to the game consoles - controlled hardware and software environments. Commodity hardware that runs a straightforward operating environment designed specifically to do common computing tasks. 32 bit processors are ideal, have smaller transistor requirements, simpler busses and simpler logic than 64 bit brothers.
I think the computer market is maturing from glitz and glammour to reliability. People are getting to the point where they realize that a 1 gigahertz computer will type letters, send email, and play dvds just as well as a 2 gHz computer.
Lots of people have tried to create a commodity computer - but, there have been few successes. In fact, the only notable ones are the iMac and iBook. Niether are powerhouses, they don't even play games. But, they're great for a distinct majority of the computer using audience. They're cheap and have tightly controlled hardware. Apple is even trying to control software, they're making quality apps to suit common computing needs. iTunes, iPhoto, etc.
I'm a big fan of Apple for their move to OSX, but I appreciate the elegance of their iMac/iBook solution. They're not making Porsche's, they're trying to make Hondas. So far, they're doing a damn good job. Some people need Porsche's and they buy them. If you need 450 horses and 490 foot pounds of torque, you can get it - but, most people don't need it.
Actually, no. The 8088/8086 are really 16 bit processors. They both had a 20 bit address bus, and had 16 bit wide registers. 8088 had an 8 bit data bus; 8086 had a 16 bit data bus. The 80286 added a primitive mmu to get the 286 protected mode and had a 24 bit address bus and 16 bit data bus. The 80386 had a 32 bit address bus, with the same old 16 bit data bus. The 386 also sported 32 bit registers and had a real mmu. The 16 bit registers, ax, bc, etc. were extended to 32 bits eax, ebx. The 16 bit commands still operated on the lower 16 bits of the registers. (register size is specified in the instruction prefix.) The 80486 had the full 32/32 compliment. (That's why you needed simms in pairs on 486's.) The 8086 to the 286 was a stumble, the 386 was definately an improvement.
However, realistically - the x86 line should have stopped around the time of the 386, 486, or perhaps the first pentiums.
The itanium is pretty disgusting. Seriously stupid flaws, for example: there is no integer division - you HAVE to use the FPU for all divisions, including transferring values to and from floating point registers.
I personally hold that Alphas are, architecturally, the king of 64 bit chips. Unfortunately, they don't have the money behind them that x86 does - I think a lot of the industry laments that. As the Pentium IV proves, with enough money you can make anything fast and cheap. Imagine what we could have done with the Alpha.
Intel doesn't care about efficiency, megahertz, or pretty much anything except... megabucks.
>I'm from Argentina: Tango, Asado, Mate, Gaucho, Maradona, Perón, Evita, el che, Borges, Pampas, Patagonia, etc.
:(
milanesas, empanadas, y sin plata.
Each context-switch also requires a context save, restore, cache-flush etc. The damage isn't too much worse than a context-switch, which happens thousands of times a second, anyways; but, it can impact maximum throughput, because the preempted task will be postponed.
With this patch the kernel becomes preemptible - meaning, other kernel tasks can stop the current one from executing, execute, finish, and allow the stopped tasks to finish.
Net effect - expensive operations can be suspended for user interactiveness. Can this impact performance, Yes. Noticeably? No.
If you're running a big-ass server, it's probably head-less, anyways - and you won't have any large, interactive processes preempting the kernel for smoothness.
If you're running a workstation, this means that X won't bog down as much when you're running those huge simulations, compiles, etc.
If you're on an embedded device, you can use this to try and get real-time responsiveness. (perhaps not ideal, but, in an embedded situation you have enough control that if you need a better real-time guarantee, you have other options (e.g. rtlinux).)
If you're on a modest, consumer PC - X won't suck as much.
All in all, this is a good idea. In theory, you lose some efficiency making several thousand context switches/second, but that's the price you pay for multi-tasking. Yeah, certain kernel operations may take longer, but, you get a better responsiveness, which - for most people, is a good thing. Most interactive individuals are seldomly pegging their processor at 100% utilization for any worthwhile period of time. (Games are an exception.)
This is good stuff.
If you're doing digital input and output, remember that the soundcard is now acting as a glorified serial port. The MAudio Delta 2448, for example, is a C-Media 8738, and the MAudio Delta 2496 is an Envy24. The EMU10K1, which powers all the versions of the live, resamples all output to 48khz - it does hurt the output. The CMI-8738, on the other hand, can output at 32khz, 44.1khz, and 48khz without a problem. The best thing is, you can purchase some C-Media cards, with decent (not midiman quality, however) drivers and optical + coxial S/PDIF in/out ports for around $40. The Midiman MAudio Delta 2448 is still around $200.
Don't necessarily avoid game cards - but, shop around and evaluate your needs. I'd trust seperate, individual ADC/DAC pairs coupled to a $40 CMI-8738 over a $200 analog soundcard by itself. (note, the ADCs/DACs themselves may cost more than $200 - but, the point is the sound card itself is secondary.)
But, most definately read what other people have to say about different cards.
In my humble quest to see why Realtek 8139C cards were $10ea, invididually, and some people still charge real money for Intel, 3Com, and Adaptec (starfire) boards, I found that some of the more quality boards had bigger buffers, and more features.
But, not meaning to knock cheap chipsets, the C-Media 8738, the heart of most $30 5.1 soundcards, has a better SPDIF implementation than the SB Live (the C-Media can generate different clock signals, rather than resampling all output to 48khz. This is a big deal, when most streams are 44.1khz) The Midiman 2448, a professional quality Coax and Optical SPDIF card - is just a C-Media 8738, with a $200 pricetag. The Midiman 2496, btw, is just a card with an Envy24 chipset.
So, yeah, I don't have the facts to back it up - but, I know that the 100 meg interfaces in the Cisco firewall products are Intel i82559s. (Open up a PIX 515 sometime - how's that for a $7,000 pent 200?)
Yeah, Intel Nics are better than realteks.