This chipset looks great if only from a technological perspective (i.e. even if they don't sell many retail, it does remind people that Matrox still exists), however the FSAA technique is amazing and will actually be used. For those who didn't read the English translated: Instead of rendering the entire screen at 4x X 4x the resolution (which is unbelievably demanding on the memory pipeline, and every part in between), instead it only oversamples polygon edges where the AA actually matters (the article mentions that this means that only 3-5% of the pixels of an average image are oversampled, dramatically reducing the demands on the system). I have a GF3 and I've never used AA because of the damands it puts on the card, but the way Matrox has done it might make it usable. We'll see I suppose.
That's why contracts are a negotiable thing, and customers either agree or disagree, and both sides decide how to proceed. There are no certainties in agreements like that, and if a company has no desire to own the IP, and the contractor can charge a lower rate (note that you can look at it two ways: You can say that handing over the IP leads to a higher rate, or you can reverse that and say that keeping the IP reduces the rate) because they can leverage it in the future, then that's what they agree on.
I use VMWare to test cross network applications on "separate" Windows boxes, all on one PC. I use it to develop applications that communicate between Windows and Linux (again, without the BS of having a tonne of PCs under my desk). I use it to test installs (the rollback feature of VMWare absolutely rocks). Swapping HDs is comletely different and is totally incomparable.
One thing about VMWare 3.x and Redhat 7.2 that I've experienced everytime I've installed it: When I start up into Gnome the resolution is 320x200: This happen to anyone else? I always get it going at my preferred 800x600 for virtual sessions, but it still is annoying and odd that it does that.
As a side note: VMWare is awesome. I love that product, and it has saved me so much time, and removed the need to have clusters of PCs around my desk. It is a product that is closest to "magic" to me (yes, I realize how it works, however it still amazes and astounds me. Especially considering that apps in the virtual OS run in the background (i.e. if they don't use GUI elements) about 85% of the speed of running direct in the base OS, and that's pretty damn impressive.
If you are referring to the Active Server Pages engine (asp.dll), and IIS, both are optional components that do not have to be installed, and can be entirely removed (Add/Remove programs, Windows Components). Furthermore, you can then go nuts and install Apache if that is your desire.
Totally agree. I was looking for the gunfight scene, or the mass explosions that would make this at all interesting or exceptional, but instead it's a "user not found" email address. The fact that this was actually noticed by the sysadmin (given the enormously small quantities of actual spam...I get more in a day than it sounds like he got in the entire span of this bogus account) either indicates someone who has too much time on their hands, or is very on top of things (perhaps too much so, which brings you to the prior possibility in a recursive fashion)
You should care about the lumens output of the bulb, not the energy consumption (i.e. W!=light output. A 15W flourescent puts out the same energy as a 60W incandescent, for example). I don't know the efficiency of the white LEDs, so I can't comment on the lumens output of it.
I'm not a fan of either Window Media player (does this POS ever actually succeed when it tries to open media in the browser? I've had it fail endlessly on several separate PCs, when the same stream opens fine in a standalone media player) or Real Player (because of exactly what you mentioned), and long ago vowed to never install RealPlayer again, but recently I did because of a local radio station that has an internet radio broadcast at a lowly 32Kbps : The Media Player feed sounds horrendous, but the Real stream at the same rate sounds very acceptable. I don't know what they did or how they did it, but at the low bit rates I'll take Real over Media Player any day.
That's pretty fringe compared to the vast swaths of people that use Limeware et. all purely to download the latest hit on the radio. There are some good sites that categorize music nicely and actually lend themselves to this musical exploration, and mp3.com would be one such example.
I can totally appreciate the idea of buying bargain PCs, and the Athlon XP 1800+ system that I bought I bought only once it got to the high value point (the chip can be bought for $189 CDN here, which is quite a sweet point). I recently bought a GeForce 3 Ti200 because it was unbelievably cost effective.
However, that doesn't man that one should yawn at new PC components coming out: I love every new graphics card, processor, and memory technology -> Maybe I'm not going to buy it today, but in 3 months it'll be bargain basement and I'll be very grateful.
How does a song that you don't know, from a band that you've never heard, expose itself to you on any of the P2P programs? The reality, of course, is that it doesn't: The only most people ever hear it is if they pulled a fast one and labelled it "Britney Spears duo with Christina Aguilara.mp3". The P2P networks are all built around searching, meaning that you hear a song on the radio, or in a movie, and you think "Hey! Remind me to download that from Gnutella later!".
Speak for yourself, and it should be noted that this exact same argument has played out since the days of the 386. I virtually guarantee you that if you search the forums, you can find dozens of people proclaiming that the new 486 is well and good, but their 386 does more than they ever need, etc. It's a dated, predictable argument.
I recently upgraded from a P3 850/512MB PC133 to an Athlon XP 1800+ with 512MB of DDR333, and the difference in simple day to day applications is stunning (one of those "You don't know better until you've tried better" type deals), and for applications like Java or Visual Studio.NET it's a requirement.
Which is why I mentioned that it could run on a couple of watts : Clearly you're not going to recover 100%, and any conversion always entails a loss, however some recovery is far better than none. It's a well known fact that the internal combustion is primarily a heat creator, rather than the torque creator that we'd prefer: Recovery for things like that would be tremendous.
Okay, so the 80% or so of a car engine's output that is wasted in heat we'll just, oh, light on fire?
The point is that there are tremendous areas of society where there is heat that's in small enough quantities that it can't be converted into energy (i.e. you can build a thermal-electric plant on top of VERY large sources of heat, but not on small sources), so any breakthrough that would allow us to recapture small scale heat (which applies to just about everything...a computer system could run on a couple of W if you could reclaim all of the heat it generates).
Much more profound though is that they're basically talking about a device that converts heat into light: The ramifications and applications of that are wide ranging and staggering. Getting even more "goofy", could you have a heat->light conversion, followed by a light->electricity conversion? (i.e. a small "heat energy recovery system").
The little incandescent replacement bulbs that are flourescents have no visible flicker whatsoever (I presume they have copious amounts of a phosphor like material on the inside equalizing between pulses), and their light spectrum, albeit a little less red, is much purer white than an incandescent. I've nothing but good things to say about them, and like the other poster I've replaced only one in a long, long time. Now that you can get them at places like IKEA for just a small price premium over incandescents, they're a no-brainer (well, unless this technology described in this article comes to the consumer level, as this describes an end result that is far more efficient than even incandescents).
I've been in quite a few meetings where the distinct, and annoying, sound of someone's buzzing phone/pager was evident for all to hear. Simply going to vibrating isn't the solution: Set the phone to silent in venues like that. As I mentioned my phone is on 100% silent mode 98% of the time, and the other 2% is when I am expecting an incoming call that I have to answer.
The mockery of individualized ring tones by people just seems, well, odd. Firstly the idea that everyone has the same individualized ring tone is silly: There are thousands of ring tones out there. But the empirical evidence is that most people haven't figured out how to switch their phone off of the Nokia songlette.
And the Cutesy Songy Ringtones get annoying after the 20th time in an hour
Just as a bit of a defence for those with "cutesy" songy ringtones : As anyone with a cellphone knows, the simplistic sine wave tones generated by todays cellphones are extremely hard to audio-locate, which is why you get the situation where 8 people all are reaching for their cell phones simultaneously. Instead I have my phone play a little song that I sent it from some website, and I know that if I hear that theme that it is overwhelmingly likely that it is my phone within the first three tones or so (versus the countless stock Nokia tone people). Mind you, 98% of the time I have my phone on silent anyways, and if I do have the ringer on it's on volume level 1 at most.
The telephone really is a fascinating part of our society: So many people are brought up believing that the telephone is instant attention from the receiving end. I personally almost never answer the telephone (that's what voicemail is for. Note that people who hate voicemail are usually the "BUT I'M TOO IMPORTANT FOR VOICEMAIL! WHERE ARE YOU! I NEED YOU UNDIVIDED ATTENTION NOW BECAUSE I'M SPECIAL AND SUPERCEDE ALL OTHER TASKS!"). Speaking of cell phones : What's with the people who always have the volume of the ringer on super-loud, and they yell into their cellphones? Totally unnecessary, and again I think it's a little too much self-importance.
Firewire was actually originally intended as a SCSI-3 serial format.
Anyways, firstly can you tell me where I can find the benchmarking program that StorageReview uses? I would love to duplicate their results to see how my system performs, and I can virtually guarantee that my CPU usage will be dramatically lower. Why? Well I noticed in their testbed docs (while searching for their benchmarking suite): "Important Note: We've stuck with Windows XP's native driver rather than installing Intel's "Application Accelerator," i.e., Intel's busmastering drivers.'". That is pure insanity. Both VIA, Intel, and the add-in maker Promise have offered high performance, low CPU usage busmastering drivers for years (drivers that can take advantage of busmastering, and the 32-command queue that most IDE controllers have nowadays, even the $20 ones), and failing to use them renders the results ridiculously irrelevant. Disingenously saying "Oh well we went with the OS drivers for the Adaptec too" is flawed as well : The IDE drivers are generic, whereas the Adaptec drivers are specific. Install the proper drivers and then come back.
Let me restate what I said : 99.99_% of the population has no need for extreme throughput, and those who do use the unbelievably expensive, extremely high throughput drive arrays. Just because someone, somewhere, does extreme HDTV video editing means absolutely, positively nothing for what I use on my development workstation. For any company to say "this product needs SCSI!" is just dumb : There are slow SCSI subsystems, and fast IDE subsystems, and vice versa. Current IDE controllers offer command queueing just like SCSI. This whole thread is almost ridiculous because it rehashes so many tired anti-IDE conclusions that were valid perhaps 7 years ago, but that no longer are true.
SCSI is used in servers because servers, almost without fail, require hot swappability, which is something that IDE wasn't intended to do (not to mention that SCSI is the "monopoly" entrenched standard in servers. Just because it's a standard doesn't mean it's superior, which is something that I'm sure most Slashdotters can agree with). I don't need that feature on my home machine, so IDE is the best choice for me, and the same holds true for most workstations (which is what we're talking about).
USB 2 and Firewire are both only about 400Mbps, so both represent significant speed downgrades from either IDE or the newer SCSIs, though they do make total sense for tape backups, scanners, etc. Personally I'm more interesting in Serial ATA as an up and coming standard.
I'm confused.. you say you have 2 drives, striped, and then talk about copying big files between them? IF they are striped they are one volume, and you can't copy things between them.
That was a mistype on my part, and what I meant to type was that while I have two 60GB drives off of a RAID controller, I haven't taken the plunge and striped them yet. As such they're both hanging as the single drive on their own bus, on two separate buses obviously.
I think the only reason IDE is more cpu intensive
It should be pointed out that while this is constantly restated, repetition doesn't count as evidence. It was ironic that just prior to seeing this debate, I saw this page which shows significantly higher CPU utilization for the two SCSI drives (mind you, they're extremely high performance drives, however they're not of a scale that would justify the difference between them and the IDEs). Each new time I replace my workstation I go through the whole IDE versus SCSI debate because I want to go with what's best (SCSI just has an air of superiority around it, much like Honda enthusiasts feel about their 115 lb-ft of torque VTEC engines : Enthusiasm, again, doesn't indicate that it's rational or based on any truths), but it seems that, firstly, it's extremely hard to find cold hard facts on the matter (i.e. basic metrics. Most of the evidence is anecdotal or based on uneven systems), but secondly that a lot of SCSI enthusiasts are very emotional about it. I have zero faith in anyone's personal opinion about the "feel" of one over the other: I remember back in the BBS days when a program made the rounds that promised to "convert your 386 to a 486!" and people would argue with me and ASSURE me that, yup, it made their system that much faster and smoother. A little persuasion and predisposition goes a long way when it comes to subjective measures, which is why I usually discount them.
Most "Joe Blow"s would indeed rather a 60GB drive over a more expensive 18GB drive when the only superiority of the 18GB drive is hypotheticals and religious zealotry. Thanks, we're all stocked up on crazy, but I think I'll use my brain and go for the superior solution (obviously IDE on workstations).
The FUD comes from the SCSI camp. SCSI is not realistic on workstations anymore, and even the much vaunted "lower CPU usage" is a claim that isn't demonstratable in real world use.
Personally I think SCSI is cheap for cheapo Joe Blows : My workstation only runs on fiber connected triple redundant NAS. I'm 31337, right?
Sounds more like the "sour grapes" ones are the fools who rushed to buy SCSI (having heard that it's "clearly superior" by fellow nutbars), only to see their friends buying vastly superior I/O systems for 1/3 the price.
As a point by point rebuttal:
1) Perhaps you have some amazing need for drives faster than 7200RPM, but the reality is that 99.99999% of the public doesn't, hence why you don't find them on the marketplace for IDE. There isn't some magical difference between IDE and SCSI that limits IDE from having faster drives, except for a marketplace that simply wouldn't be willing to pay for it (the "non-nutbar" community).
2) See point 1.
3) I have _never_ had a need for more drives, though my current motherboard supports up to 8 IDE devices, right now I have two CDs, and two HDs, and with 120GB I'm not begging to stick another 18GB "superior" drive in to show my expansion capabilities. For those who say "but SCSI lets you connect tape drives/scanners/whatever", realize that 1394 or USB 2 are far better solutions for that anyways (and my MB has USB 2.0 and Firewire on it).
4) Prove it. This old BS has been around since the days of IDE controllers that didn't support busmastering, and the same regurgitated nonsense continually reappears. Again, drop the anecdotal BS and put together a comparison of a grossly overpriced Adaptec controller with a sucker-punch SCSI drive, and a ATA-133 with a current drive : I doubt CPU usage will vary more than 1%.
But, oh boy I wish I could spend lots of money on an equal or lower performing SCSI system: That's my dream! Oh boy I can only hope that one day I'll achieve that goal.
Why don't you put together a benchmark then and document your methodology? The reality is that this SCSI dream is a religion, not a reality. SCSI seems to be sacred grounds for many on here, usually for anecdotal reasons. Bring on the metrics and I'm willing to be convinced in a moment, but for desktop/workstation use (note that I differentiate servers), SCSI usually indicates the naivety of the purchaser more than the performance of the system.
I have a pretty good IDE controller (the Promise RAID Lite on a Asus A7V333 motherboard), with a pair of fairly fast hard drive (Maxtor Diamondmax 7200RPM 60GB. Configured as two separate drives [I have taken the plunge and RAID 0 striped them]) and when I copy gigantic, multi-hundred MB files between them the CPU usage hovers around a blistering 0%.
Really I think the "coprocessing" nature of SCSI is grossly, overwhelmingly overstated. Perhaps when CPUs were piddly little 386s this made a huge difference, but the logic required to manage the disks on something like a P4 or an Athlon XP is so unbelievably trivially (not to mention that most IDE controllers offload anyways, and all IDE implementations for about the past 8 years have supported busmastering) low, I really question the veracity of the claims about SCSI's advantage in that area. Even if SCSI did reduce my CPU usage 1% (I doubt even that), I think I'll still go with ATA-133 and buy 3x bigger drives for 1/3 the price.
Slashdot Mirror
This chipset looks great if only from a technological perspective (i.e. even if they don't sell many retail, it does remind people that Matrox still exists), however the FSAA technique is amazing and will actually be used. For those who didn't read the English translated: Instead of rendering the entire screen at 4x X 4x the resolution (which is unbelievably demanding on the memory pipeline, and every part in between), instead it only oversamples polygon edges where the AA actually matters (the article mentions that this means that only 3-5% of the pixels of an average image are oversampled, dramatically reducing the demands on the system). I have a GF3 and I've never used AA because of the damands it puts on the card, but the way Matrox has done it might make it usable. We'll see I suppose.
That's why contracts are a negotiable thing, and customers either agree or disagree, and both sides decide how to proceed. There are no certainties in agreements like that, and if a company has no desire to own the IP, and the contractor can charge a lower rate (note that you can look at it two ways: You can say that handing over the IP leads to a higher rate, or you can reverse that and say that keeping the IP reduces the rate) because they can leverage it in the future, then that's what they agree on.
I use VMWare to test cross network applications on "separate" Windows boxes, all on one PC. I use it to develop applications that communicate between Windows and Linux (again, without the BS of having a tonne of PCs under my desk). I use it to test installs (the rollback feature of VMWare absolutely rocks). Swapping HDs is comletely different and is totally incomparable.
VMware rocks.
One thing about VMWare 3.x and Redhat 7.2 that I've experienced everytime I've installed it: When I start up into Gnome the resolution is 320x200: This happen to anyone else? I always get it going at my preferred 800x600 for virtual sessions, but it still is annoying and odd that it does that.
As a side note: VMWare is awesome. I love that product, and it has saved me so much time, and removed the need to have clusters of PCs around my desk. It is a product that is closest to "magic" to me (yes, I realize how it works, however it still amazes and astounds me. Especially considering that apps in the virtual OS run in the background (i.e. if they don't use GUI elements) about 85% of the speed of running direct in the base OS, and that's pretty damn impressive.
If you are referring to the Active Server Pages engine (asp.dll), and IIS, both are optional components that do not have to be installed, and can be entirely removed (Add/Remove programs, Windows Components). Furthermore, you can then go nuts and install Apache if that is your desire.
Totally agree. I was looking for the gunfight scene, or the mass explosions that would make this at all interesting or exceptional, but instead it's a "user not found" email address. The fact that this was actually noticed by the sysadmin (given the enormously small quantities of actual spam...I get more in a day than it sounds like he got in the entire span of this bogus account) either indicates someone who has too much time on their hands, or is very on top of things (perhaps too much so, which brings you to the prior possibility in a recursive fashion)
You should care about the lumens output of the bulb, not the energy consumption (i.e. W!=light output. A 15W flourescent puts out the same energy as a 60W incandescent, for example). I don't know the efficiency of the white LEDs, so I can't comment on the lumens output of it.
I'm not a fan of either Window Media player (does this POS ever actually succeed when it tries to open media in the browser? I've had it fail endlessly on several separate PCs, when the same stream opens fine in a standalone media player) or Real Player (because of exactly what you mentioned), and long ago vowed to never install RealPlayer again, but recently I did because of a local radio station that has an internet radio broadcast at a lowly 32Kbps : The Media Player feed sounds horrendous, but the Real stream at the same rate sounds very acceptable. I don't know what they did or how they did it, but at the low bit rates I'll take Real over Media Player any day.
That's pretty fringe compared to the vast swaths of people that use Limeware et. all purely to download the latest hit on the radio. There are some good sites that categorize music nicely and actually lend themselves to this musical exploration, and mp3.com would be one such example.
I can totally appreciate the idea of buying bargain PCs, and the Athlon XP 1800+ system that I bought I bought only once it got to the high value point (the chip can be bought for $189 CDN here, which is quite a sweet point). I recently bought a GeForce 3 Ti200 because it was unbelievably cost effective.
However, that doesn't man that one should yawn at new PC components coming out: I love every new graphics card, processor, and memory technology -> Maybe I'm not going to buy it today, but in 3 months it'll be bargain basement and I'll be very grateful.
How does a song that you don't know, from a band that you've never heard, expose itself to you on any of the P2P programs? The reality, of course, is that it doesn't: The only most people ever hear it is if they pulled a fast one and labelled it "Britney Spears duo with Christina Aguilara.mp3". The P2P networks are all built around searching, meaning that you hear a song on the radio, or in a movie, and you think "Hey! Remind me to download that from Gnutella later!".
Speak for yourself, and it should be noted that this exact same argument has played out since the days of the 386. I virtually guarantee you that if you search the forums, you can find dozens of people proclaiming that the new 486 is well and good, but their 386 does more than they ever need, etc. It's a dated, predictable argument.
I recently upgraded from a P3 850/512MB PC133 to an Athlon XP 1800+ with 512MB of DDR333, and the difference in simple day to day applications is stunning (one of those "You don't know better until you've tried better" type deals), and for applications like Java or Visual Studio.NET it's a requirement.
Which is why I mentioned that it could run on a couple of watts : Clearly you're not going to recover 100%, and any conversion always entails a loss, however some recovery is far better than none. It's a well known fact that the internal combustion is primarily a heat creator, rather than the torque creator that we'd prefer: Recovery for things like that would be tremendous.
Okay, so the 80% or so of a car engine's output that is wasted in heat we'll just, oh, light on fire?
The point is that there are tremendous areas of society where there is heat that's in small enough quantities that it can't be converted into energy (i.e. you can build a thermal-electric plant on top of VERY large sources of heat, but not on small sources), so any breakthrough that would allow us to recapture small scale heat (which applies to just about everything...a computer system could run on a couple of W if you could reclaim all of the heat it generates).
Much more profound though is that they're basically talking about a device that converts heat into light: The ramifications and applications of that are wide ranging and staggering. Getting even more "goofy", could you have a heat->light conversion, followed by a light->electricity conversion? (i.e. a small "heat energy recovery system").
The little incandescent replacement bulbs that are flourescents have no visible flicker whatsoever (I presume they have copious amounts of a phosphor like material on the inside equalizing between pulses), and their light spectrum, albeit a little less red, is much purer white than an incandescent. I've nothing but good things to say about them, and like the other poster I've replaced only one in a long, long time. Now that you can get them at places like IKEA for just a small price premium over incandescents, they're a no-brainer (well, unless this technology described in this article comes to the consumer level, as this describes an end result that is far more efficient than even incandescents).
I've been in quite a few meetings where the distinct, and annoying, sound of someone's buzzing phone/pager was evident for all to hear. Simply going to vibrating isn't the solution: Set the phone to silent in venues like that. As I mentioned my phone is on 100% silent mode 98% of the time, and the other 2% is when I am expecting an incoming call that I have to answer.
The mockery of individualized ring tones by people just seems, well, odd. Firstly the idea that everyone has the same individualized ring tone is silly: There are thousands of ring tones out there. But the empirical evidence is that most people haven't figured out how to switch their phone off of the Nokia songlette.
And the Cutesy Songy Ringtones get annoying after the 20th time in an hour
Just as a bit of a defence for those with "cutesy" songy ringtones : As anyone with a cellphone knows, the simplistic sine wave tones generated by todays cellphones are extremely hard to audio-locate, which is why you get the situation where 8 people all are reaching for their cell phones simultaneously. Instead I have my phone play a little song that I sent it from some website, and I know that if I hear that theme that it is overwhelmingly likely that it is my phone within the first three tones or so (versus the countless stock Nokia tone people). Mind you, 98% of the time I have my phone on silent anyways, and if I do have the ringer on it's on volume level 1 at most.
The telephone really is a fascinating part of our society: So many people are brought up believing that the telephone is instant attention from the receiving end. I personally almost never answer the telephone (that's what voicemail is for. Note that people who hate voicemail are usually the "BUT I'M TOO IMPORTANT FOR VOICEMAIL! WHERE ARE YOU! I NEED YOU UNDIVIDED ATTENTION NOW BECAUSE I'M SPECIAL AND SUPERCEDE ALL OTHER TASKS!"). Speaking of cell phones : What's with the people who always have the volume of the ringer on super-loud, and they yell into their cellphones? Totally unnecessary, and again I think it's a little too much self-importance.
Firewire was actually originally intended as a SCSI-3 serial format.
Anyways, firstly can you tell me where I can find the benchmarking program that StorageReview uses? I would love to duplicate their results to see how my system performs, and I can virtually guarantee that my CPU usage will be dramatically lower. Why? Well I noticed in their testbed docs (while searching for their benchmarking suite): "Important Note: We've stuck with Windows XP's native driver rather than installing Intel's "Application Accelerator," i.e., Intel's busmastering drivers.'". That is pure insanity. Both VIA, Intel, and the add-in maker Promise have offered high performance, low CPU usage busmastering drivers for years (drivers that can take advantage of busmastering, and the 32-command queue that most IDE controllers have nowadays, even the $20 ones), and failing to use them renders the results ridiculously irrelevant. Disingenously saying "Oh well we went with the OS drivers for the Adaptec too" is flawed as well : The IDE drivers are generic, whereas the Adaptec drivers are specific. Install the proper drivers and then come back.
Let me restate what I said : 99.99_% of the population has no need for extreme throughput, and those who do use the unbelievably expensive, extremely high throughput drive arrays. Just because someone, somewhere, does extreme HDTV video editing means absolutely, positively nothing for what I use on my development workstation. For any company to say "this product needs SCSI!" is just dumb : There are slow SCSI subsystems, and fast IDE subsystems, and vice versa. Current IDE controllers offer command queueing just like SCSI. This whole thread is almost ridiculous because it rehashes so many tired anti-IDE conclusions that were valid perhaps 7 years ago, but that no longer are true.
SCSI is used in servers because servers, almost without fail, require hot swappability, which is something that IDE wasn't intended to do (not to mention that SCSI is the "monopoly" entrenched standard in servers. Just because it's a standard doesn't mean it's superior, which is something that I'm sure most Slashdotters can agree with). I don't need that feature on my home machine, so IDE is the best choice for me, and the same holds true for most workstations (which is what we're talking about).
USB 2 and Firewire are both only about 400Mbps, so both represent significant speed downgrades from either IDE or the newer SCSIs, though they do make total sense for tape backups, scanners, etc. Personally I'm more interesting in Serial ATA as an up and coming standard.
I'm confused.. you say you have 2 drives, striped, and then talk about copying big files between them? IF they are striped they are one volume, and you can't copy things between them.
That was a mistype on my part, and what I meant to type was that while I have two 60GB drives off of a RAID controller, I haven't taken the plunge and striped them yet. As such they're both hanging as the single drive on their own bus, on two separate buses obviously.
I think the only reason IDE is more cpu intensive
It should be pointed out that while this is constantly restated, repetition doesn't count as evidence. It was ironic that just prior to seeing this debate, I saw this page which shows significantly higher CPU utilization for the two SCSI drives (mind you, they're extremely high performance drives, however they're not of a scale that would justify the difference between them and the IDEs). Each new time I replace my workstation I go through the whole IDE versus SCSI debate because I want to go with what's best (SCSI just has an air of superiority around it, much like Honda enthusiasts feel about their 115 lb-ft of torque VTEC engines : Enthusiasm, again, doesn't indicate that it's rational or based on any truths), but it seems that, firstly, it's extremely hard to find cold hard facts on the matter (i.e. basic metrics. Most of the evidence is anecdotal or based on uneven systems), but secondly that a lot of SCSI enthusiasts are very emotional about it. I have zero faith in anyone's personal opinion about the "feel" of one over the other: I remember back in the BBS days when a program made the rounds that promised to "convert your 386 to a 486!" and people would argue with me and ASSURE me that, yup, it made their system that much faster and smoother. A little persuasion and predisposition goes a long way when it comes to subjective measures, which is why I usually discount them.
Most "Joe Blow"s would indeed rather a 60GB drive over a more expensive 18GB drive when the only superiority of the 18GB drive is hypotheticals and religious zealotry. Thanks, we're all stocked up on crazy, but I think I'll use my brain and go for the superior solution (obviously IDE on workstations).
The FUD comes from the SCSI camp. SCSI is not realistic on workstations anymore, and even the much vaunted "lower CPU usage" is a claim that isn't demonstratable in real world use.
Personally I think SCSI is cheap for cheapo Joe Blows : My workstation only runs on fiber connected triple redundant NAS. I'm 31337, right?
Sounds more like the "sour grapes" ones are the fools who rushed to buy SCSI (having heard that it's "clearly superior" by fellow nutbars), only to see their friends buying vastly superior I/O systems for 1/3 the price.
As a point by point rebuttal:
1) Perhaps you have some amazing need for drives faster than 7200RPM, but the reality is that 99.99999% of the public doesn't, hence why you don't find them on the marketplace for IDE. There isn't some magical difference between IDE and SCSI that limits IDE from having faster drives, except for a marketplace that simply wouldn't be willing to pay for it (the "non-nutbar" community).
2) See point 1.
3) I have _never_ had a need for more drives, though my current motherboard supports up to 8 IDE devices, right now I have two CDs, and two HDs, and with 120GB I'm not begging to stick another 18GB "superior" drive in to show my expansion capabilities. For those who say "but SCSI lets you connect tape drives/scanners/whatever", realize that 1394 or USB 2 are far better solutions for that anyways (and my MB has USB 2.0 and Firewire on it).
4) Prove it. This old BS has been around since the days of IDE controllers that didn't support busmastering, and the same regurgitated nonsense continually reappears. Again, drop the anecdotal BS and put together a comparison of a grossly overpriced Adaptec controller with a sucker-punch SCSI drive, and a ATA-133 with a current drive : I doubt CPU usage will vary more than 1%.
But, oh boy I wish I could spend lots of money on an equal or lower performing SCSI system: That's my dream! Oh boy I can only hope that one day I'll achieve that goal.
Why don't you put together a benchmark then and document your methodology? The reality is that this SCSI dream is a religion, not a reality. SCSI seems to be sacred grounds for many on here, usually for anecdotal reasons. Bring on the metrics and I'm willing to be convinced in a moment, but for desktop/workstation use (note that I differentiate servers), SCSI usually indicates the naivety of the purchaser more than the performance of the system.
I have a pretty good IDE controller (the Promise RAID Lite on a Asus A7V333 motherboard), with a pair of fairly fast hard drive (Maxtor Diamondmax 7200RPM 60GB. Configured as two separate drives [I have taken the plunge and RAID 0 striped them]) and when I copy gigantic, multi-hundred MB files between them the CPU usage hovers around a blistering 0%.
Really I think the "coprocessing" nature of SCSI is grossly, overwhelmingly overstated. Perhaps when CPUs were piddly little 386s this made a huge difference, but the logic required to manage the disks on something like a P4 or an Athlon XP is so unbelievably trivially (not to mention that most IDE controllers offload anyways, and all IDE implementations for about the past 8 years have supported busmastering) low, I really question the veracity of the claims about SCSI's advantage in that area. Even if SCSI did reduce my CPU usage 1% (I doubt even that), I think I'll still go with ATA-133 and buy 3x bigger drives for 1/3 the price.