If you're using RAID 0 for your swap, then there's no reason to even use RAID. Simply create two swap partitions of equal priority, and the kernel will automatically split the load between them.
(Now if you were using a redundant form of RAID, that's a different story. No operating system likes having the swap pulled out from under it, so keeping it alive through a disk failure is a good thing.)
As has been the case for a long, long time, the cabling isn't going to be the limitting factor, it'll be the routing.
Sure, we can wire up a 300-house neighborhood with 100 mbit fiber connections. But there's a potential for 30 gigabits/second of traffic in the pathlological case. Switching 30 gigabits/second can be done, but routing 30 gigabits is another story - and it'll have to be routed several times to get it along and out of your network.
Let's say that a medium-sized city with 100,000 homes each got a 10 megabit connection. The cost of the routers to handle that kind of potential load would greatly eclipse the cost of the network itself.
So, what would the company do? Keep prices high for bandwidth. That way, the installed customer-base doesn't get so large that it's impossible, or even economically unfeasable to support it. As time goes on, router performance goes up, and (if you're lucky) competitors enter the market, prices will go down.
Actually, I've seen some rewiring. Our local phone company hasn't laid copper in a very long time. There are very large parts of the city that can't get DSL precisely because the only copper is from their house to the street's phone box, from which point it's all fiber.
And while you're accusing ComCast, I spent some time last summer planting and grooming the most amazingly beautiful park strip. Then ComCast came in and re-cabled the entire neighborhood to use fiber instead of copper, digging up my park strip about 9 times in the process.
There are lots of people recabling to use fiber. It's just not always announced on SlashDot. : )
You have to be careful with that. Remember, most homes already have a connection that could make 100 mbit look like child's play: A cable television connection. There's an awfully large amount of bandwidth there, it's just used for something other than data.
Getting a 100 (or 1000) mbit connection into your home doesn't mean that you'll get a 100-mbit connection to the Internet. It just means that you *can* get whatever connection to the Internet you want, and that you can also get phone, video, and perhaps other services over the same connection.
Just because a particular technology *didn't* make money in the large scale doesn't mean that it doesn't for other companies. To use your example of VoIP, there are a number of companies making a good living at it. They're just not approaching it the same way the failed companies did.
The largest problem is that investors want companies to jump on bandwagons - to do what they thing everyone else is doing, instead of doing what will actually make sense.
If fourteen companies are spending $50 million each chasing the same pipe-dream, it's tough to tell investors that going a different route would be better. They just don't see things like that.
A 200 MHz processer, 64 megs of ram, and 32 megs of flash. For $400.
For the same price, I could put together an AthlonXP 2500+ with 512 megs of memory, a real hard drive, and a cd-burner.
I know, there are vast differences in manufacturing strategies, supply and demand, and all other items that dictate how much these handhelds will cost. But it's still a kick in the pants when a very low-powered handheld costs as much as a fairly powerful workstation.
One more thing... whatever gave you the idea that diskless terminals were stuck on an 80x25 screen?
If you're thinking of a "dumb terminal", that's one thing - but "diskless" and "dumb" are two different things. There's nothing stopping you from booting into your full-fledged GUI on a diskless terminal.
Look more into the client-server model of X-windows. You'll find that you can even run your GUI very nicely on a relatively dumb terminal
Shoot, if these things are cheap enough, I've already given a good bit of thought to buying several to use as diskless terminals in my home. I could always buy some $50 clunkers from the thrift store, but they're big, noisy, and use a lot of power. These would be much cooler, in several different senses.
I once had the DMA controller on a motherboard go bad on me. Do you think that I've stopped buying motherboards without replaceable DMA controllers?
Just because you once blew out the onboard video doesn't mean that every motherboard will have that happen, or even any more than a very few motherboards.
Besides, you're just as likely to blow a regular video card as you are to blow the VGA on this board - and that regular video card might just cost *more* to replace than this entire mortherboard!
If anything, I've found most computer hardware to be much more resiliant and hard to "blow" than I would have imagined. I've hot-(un)plugged just about every type of PC interface there is without damaging the computer (sometimes on purpose, sometimes on accident). And if I went into the stories of things I've seen people do without damaging a computer, your eyes would probably pop out of your head.
The only thing I've had damaged by hot-(un)plugging was one particular model of monitor from one particular vendor. They weren't designed well, and they'd go "pop" quite often if you plugged them in to a running computer. However, that hasn't stopped me from doing it with some uncountable number of monitors, and no others have ever given me any problem.
Besides, don't tell me that you'd never buy an Opteron for fear that the memory controller in it would get blown, rendering the rest of the CPU useless....:)
Hopefully the technological divide will dimish before a major financial catastrophe occurs
Not a chance. Until something big happens several times, there simply won't be enough of a drive to make anything better.
So many people are so content with the crap that Redmond pumps out, it's just disgusting. They're also afraid of the effort to learn anything new. Every time someone complains about popup ads, I tell them that there are other browsers they can use which will block them. Guess how many have switched! That's right, very, very few. Most people just go back to the "comfort zone".
I've met some people like that in my life. All of them suffering from very severe psychological problems. Old McBride is sounding more and more like he fits into that particular segment of society.
(Of course, living in Provo is, in and of itself, a warning sign.)
Out of the 100 or so CPU's that I've bought over the last ten years, I've had exactly *one* go bad on me. (FWIW, that was a Cyrix.)
Now, I've received some bad ones - but we're talking about a working CPU turning into a non-working CPU. That's actually a pretty rare thing - not just in my experience, but also in the experiences of my associates who have dealt in much larger numbers of CPU's than I have.
If you're had that many of your chips go bad, either (a) YOU are doing something wrong, (b) you're buying from someone involved with something shady, (c) there's something extremely abnormal in your environment, or (d) you're an amazing statistical anomoly, and should go out and buy a lottery ticket immediately.:)
That's actually an incredibly useful idea - and 99% of the technology is already out there. Mini-ITX systems with AGP/PCI slots could be built very light if the cases were done of plastic (like laptops), they're small, and 12-volt DC-DC converters already exist to power them from batteries.
The only part that doesn't exist is the unified pointer/keyboard/LCD - and that wouldn't be hard to make, either.
(Shoot, for the real size/weight-conscious, there are SBC's that are amazingly tiny and light!)
Keyboards on laptops suck, because of size constraints. Screens have sucked, because of size and power constraints. CPU performance has sucked because of power constraints. Power has sucked because of weight and size constraints.
Well, why do the size and weight have to be constrained? While there are probably a lot of people who won't buy anything over six pounds, that's not a concern to me, or to a number of other people. I'd much rather lug around a 16-pound (or even 20-pound!) notebook, if it had a good keyboard, decent performance, and decent battery life.
It really doesn't do me any good to get a 5-pound computer somewhere if I can't type on it when I get there.
Anandtech seems to do all of their business with Microsoft products, they don't really seem to be completely aware (or competent) with Unix or any Unix-like operating system. Instead of malice, I'd suggest ignorance as a more probably cause. : )
Actually, I really imagine that the benefits of an Opteron can only get better on better operating systems.
M$' garbage is made for the traditional shared-bus style of design, which the Xeon uses, and isn't designed to take full advantage of the NUMA architecture of the Opterons. With a better OS, the Opteron is likely to take even more of a lead.
M$' garbage *does* have support for hyperthreading, giving the Xeons a boost. However, it does *not* appear that the version they used had support to run in 64-bit mode on the Opteron. Even though the bitness isn't going to speed it up, 64-bit mode has more registers, which would benefit the opteron in these results. Again, with a better OS, the Opteron would be likely to take a bigger lead.
So, the benchmarks aren't perfect by any means, but they're not completely useless, either. If nothing else, it's a good way to rub Intel's nose in the fact that their overpriced chips are being beaten by bargain-priced underdogs, even when the cards are stacked in Intel's favor.
I highly doubt that HP is making their own chipsets, or coming up with any other improvements that would give them a significant *hardware* advantage over the others. And with the Opteron being so self-contained, it would be even harder. With the memory controller being integrated, as long as you have the HyperTransport lines wired up, then most of the performance difference will come from memory timings and the like.
Out of the quad-Opteron arena, the Tyan is (as you imply) a complete joke, its memory design is completely, retardedly crippled. However, the vast majority of quad-Opteron boards are either AMD's design or NewIsys' design, and those two companies are betting the farm, so to speak, on the Opteron. I have a hard time believing that either of them are just lazily selling unoptimized boards.
Also, if anything, "vanilla" boards tend to be *overly* optimized, at the possible expense of stability. As an example, if you take a look at the *actual* FSB on a good number of boards, you'll find it measurably higher than what the BIOS screen told you it was.
Yes, the tests weren't exactly apples-to-apples - the outcomes are actually much better for AMD than the graphs would initially appear.
The graphs mean that Opterons with a "measly" 1 meg of cache are beating out Xeons that have (a) four times the cache, (b) 50% higher clock speed, and (c) a price tag that's three times higher.
Hats off to AMD. In times past (K2/K3), price was the only thing they had better than Intel. Now they've got both price and performance.
I once watched an interview with a specialist in airport security. He talked about seeing a family in one of the terminals - mother, father, kids - taking a video of themselves in the terminal before a flight. No big deal - except they showed up and did it again the next day. And the next day. And the next day - for an entire week.
What were they doing? The video camera wasn't focused on the smiling, waving family - it was zoomed to the pay phones some distance behind them, recording people punching in their credit-card numbers.
This really doesn't surprise me. BSD's tend to be very secure by default, when the most popular Linux distro (RedHat) seems to strive for insecure defaults, and some other distros aren't much better. The problem isn't in Linux, it's the insecure packages that the distributors bundle up and turn on by default.
Really. If Linux distributions would turn services off by default, and leave it up to the user to turn them on, I honestly believe that the succesful attacks against Linux would be cut by at least 50%. And if they'd get away from the "classically insecure" daemons, I believe you'd cut the rate much, much farther.
While people using cheats do exist, they're not nearly as prevalent as you seem to think, you're just unable to accept that people are that good.
Take Quake 3, for example. I got tired of having my butt handed to me by people with the railgun. So, I limitted myself to *only* using the railgun. In time, I got pretty good. With more time, I got really good. With even more time, I got fan-freaking-tastic. When you're single, don't own a home, and have a 1-megabit pipe into your apartment, it's amazing how much time you can dedicate/waste on those things!
So, what did that get me? A bunch of whining cry-babies yelling "Aimbot! Cheater!" every time I killed them. After they did it enough to annoy me, I'd chase them down and kill them with the gauntlet. At that point, they'd generally just leave the game.
Does this mean they'll actually MD5 the root password?
(Sarcasm-less explanation: During the RedHat installation procedure, the ability to choose to use MD5-encrpted passwords comes *after* you choose your root password, so your root password is encrypted with much weaker encryption until you change it.)
and a lot of the routing logic is hardware based so they are able to scale & perform much better than the latter
That's only true if you're not doing anything but routing. Start adding in any features that are remotely useful, and you've gone from CEF/fast switching to process switching. And that means you take a 5x to 10x performance hit.
You actually have to purchase a *VERY* expensive Cisco to get one that can't be out-performed by even a relatively modest PC.
And even on the high-end, there are PC's that will completely blow away any but the very largest offerings from Cisco.
There are, of course, several real advantages to using a Cisco router instead of a PC. First and foremost, if you have the money for it, you're going to be able to find a network interface for almost any type of network you can imagine. DSL to SONET, the interfaces are there for the buying. I'd *almost* say that they would "just work", but that's not always the case. I just had to upgrade the IOS versio on one of my Ciscos (a *paid* upgrade) to get nothing but support for an additional ethernet WIC.
Another large (perhaps HUGE) advantage is the fact that as long as you want to keep paying the money for it, there will always be someone there to back you up and make sure your problem gets resolved, even if you're not capable of fixing it yourself.
So, I'm not saying that there's no use for Cisco. I'm just saying that absolute performance (and especially price-for-performance) are not real advantages for them.
I don't know why they would want to hold back Linksys development though
You can't? Let's think about it: Do you think they'd rather sell a $200 Linksys router, or a $2,000+ Cisco router?
I know, it sounds cynical. Unfortunately, I've worked in enough corporations to know that is *exactly* how decisions are made on these things.
They'll keep the Linksys line around for the low-end market, the home users that don't demand much. But for anything above that, they're never going to let Linksys compete with their Cisco line. They're not stupid.
The performance increase comes from a combination of lower memory latency (built-in memory controller) and an increased number of registers. The small number of registers on x86 chips has always been one of the main gripes people have had about the architecture.
Maybe because spending millions of dollars on a space mission entails rigorously testing every piece of equipment which could take years before the launch?
Yes, I'm sure that it would have taken so long to test a larger CCD and a longer lens....
an extra-thousand bucks will probably get you nowhere
So, an extra twenty thousand bucks. That's still a completely insignificant drop in the bucked in relation to the rest of the mission. What, half a million just to press the "launch" button, along with six or ten years of paying hundreds (or thousands?) of people's salaries? Twenty grand wouldn't even be noticed.
Slashdot Mirror
If you're using RAID 0 for your swap, then there's no reason to even use RAID. Simply create two swap partitions of equal priority, and the kernel will automatically split the load between them.
(Now if you were using a redundant form of RAID, that's a different story. No operating system likes having the swap pulled out from under it, so keeping it alive through a disk failure is a good thing.)
steve
As has been the case for a long, long time, the cabling isn't going to be the limitting factor, it'll be the routing.
Sure, we can wire up a 300-house neighborhood with 100 mbit fiber connections. But there's a potential for 30 gigabits/second of traffic in the pathlological case. Switching 30 gigabits/second can be done, but routing 30 gigabits is another story - and it'll have to be routed several times to get it along and out of your network.
Let's say that a medium-sized city with 100,000 homes each got a 10 megabit connection. The cost of the routers to handle that kind of potential load would greatly eclipse the cost of the network itself.
So, what would the company do? Keep prices high for bandwidth. That way, the installed customer-base doesn't get so large that it's impossible, or even economically unfeasable to support it. As time goes on, router performance goes up, and (if you're lucky) competitors enter the market, prices will go down.
steve
Actually, I've seen some rewiring. Our local phone company hasn't laid copper in a very long time. There are very large parts of the city that can't get DSL precisely because the only copper is from their house to the street's phone box, from which point it's all fiber.
And while you're accusing ComCast, I spent some time last summer planting and grooming the most amazingly beautiful park strip. Then ComCast came in and re-cabled the entire neighborhood to use fiber instead of copper, digging up my park strip about 9 times in the process.
There are lots of people recabling to use fiber. It's just not always announced on SlashDot. : )
steve
You have to be careful with that. Remember, most homes already have a connection that could make 100 mbit look like child's play: A cable television connection. There's an awfully large amount of bandwidth there, it's just used for something other than data.
Getting a 100 (or 1000) mbit connection into your home doesn't mean that you'll get a 100-mbit connection to the Internet. It just means that you *can* get whatever connection to the Internet you want, and that you can also get phone, video, and perhaps other services over the same connection.
steve
Just because a particular technology *didn't* make money in the large scale doesn't mean that it doesn't for other companies. To use your example of VoIP, there are a number of companies making a good living at it. They're just not approaching it the same way the failed companies did.
The largest problem is that investors want companies to jump on bandwagons - to do what they thing everyone else is doing, instead of doing what will actually make sense.
If fourteen companies are spending $50 million each chasing the same pipe-dream, it's tough to tell investors that going a different route would be better. They just don't see things like that.
steve
A 200 MHz processer, 64 megs of ram, and 32 megs of flash. For $400.
For the same price, I could put together an AthlonXP 2500+ with 512 megs of memory, a real hard drive, and a cd-burner.
I know, there are vast differences in manufacturing strategies, supply and demand, and all other items that dictate how much these handhelds will cost. But it's still a kick in the pants when a very low-powered handheld costs as much as a fairly powerful workstation.
steve
One more thing... whatever gave you the idea that diskless terminals were stuck on an 80x25 screen?
If you're thinking of a "dumb terminal", that's one thing - but "diskless" and "dumb" are two different things. There's nothing stopping you from booting into your full-fledged GUI on a diskless terminal.
Look more into the client-server model of X-windows. You'll find that you can even run your GUI very nicely on a relatively dumb terminal
Shoot, if these things are cheap enough, I've already given a good bit of thought to buying several to use as diskless terminals in my home. I could always buy some $50 clunkers from the thrift store, but they're big, noisy, and use a lot of power. These would be much cooler, in several different senses.
steve
I once had the DMA controller on a motherboard go bad on me. Do you think that I've stopped buying motherboards without replaceable DMA controllers?
:)
Just because you once blew out the onboard video doesn't mean that every motherboard will have that happen, or even any more than a very few motherboards.
Besides, you're just as likely to blow a regular video card as you are to blow the VGA on this board - and that regular video card might just cost *more* to replace than this entire mortherboard!
If anything, I've found most computer hardware to be much more resiliant and hard to "blow" than I would have imagined. I've hot-(un)plugged just about every type of PC interface there is without damaging the computer (sometimes on purpose, sometimes on accident). And if I went into the stories of things I've seen people do without damaging a computer, your eyes would probably pop out of your head.
The only thing I've had damaged by hot-(un)plugging was one particular model of monitor from one particular vendor. They weren't designed well, and they'd go "pop" quite often if you plugged them in to a running computer. However, that hasn't stopped me from doing it with some uncountable number of monitors, and no others have ever given me any problem.
Besides, don't tell me that you'd never buy an Opteron for fear that the memory controller in it would get blown, rendering the rest of the CPU useless....
steve
Hopefully the technological divide will dimish before a major financial catastrophe occurs
Not a chance. Until something big happens several times, there simply won't be enough of a drive to make anything better.
So many people are so content with the crap that Redmond pumps out, it's just disgusting. They're also afraid of the effort to learn anything new. Every time someone complains about popup ads, I tell them that there are other browsers they can use which will block them. Guess how many have switched! That's right, very, very few. Most people just go back to the "comfort zone".
steve
"Linux has our technology."
"We can't show proof, it would harm us."
"My enemies are out to kill me."
I've met some people like that in my life. All of them suffering from very severe psychological problems. Old McBride is sounding more and more like he fits into that particular segment of society.
(Of course, living in Provo is, in and of itself, a warning sign.)
steve
Out of the 100 or so CPU's that I've bought over the last ten years, I've had exactly *one* go bad on me. (FWIW, that was a Cyrix.)
:)
Now, I've received some bad ones - but we're talking about a working CPU turning into a non-working CPU. That's actually a pretty rare thing - not just in my experience, but also in the experiences of my associates who have dealt in much larger numbers of CPU's than I have.
If you're had that many of your chips go bad, either (a) YOU are doing something wrong, (b) you're buying from someone involved with something shady, (c) there's something extremely abnormal in your environment, or (d) you're an amazing statistical anomoly, and should go out and buy a lottery ticket immediately.
steve
That's actually an incredibly useful idea - and 99% of the technology is already out there. Mini-ITX systems with AGP/PCI slots could be built very light if the cases were done of plastic (like laptops), they're small, and 12-volt DC-DC converters already exist to power them from batteries.
The only part that doesn't exist is the unified pointer/keyboard/LCD - and that wouldn't be hard to make, either.
(Shoot, for the real size/weight-conscious, there are SBC's that are amazingly tiny and light!)
steve
Keyboards on laptops suck, because of size constraints. Screens have sucked, because of size and power constraints. CPU performance has sucked because of power constraints. Power has sucked because of weight and size constraints.
Well, why do the size and weight have to be constrained? While there are probably a lot of people who won't buy anything over six pounds, that's not a concern to me, or to a number of other people. I'd much rather lug around a 16-pound (or even 20-pound!) notebook, if it had a good keyboard, decent performance, and decent battery life.
It really doesn't do me any good to get a 5-pound computer somewhere if I can't type on it when I get there.
steve
Anandtech seems to do all of their business with Microsoft products, they don't really seem to be completely aware (or competent) with Unix or any Unix-like operating system. Instead of malice, I'd suggest ignorance as a more probably cause. : )
steve
Actually, I really imagine that the benefits of an Opteron can only get better on better operating systems.
M$' garbage is made for the traditional shared-bus style of design, which the Xeon uses, and isn't designed to take full advantage of the NUMA architecture of the Opterons. With a better OS, the Opteron is likely to take even more of a lead.
M$' garbage *does* have support for hyperthreading, giving the Xeons a boost. However, it does *not* appear that the version they used had support to run in 64-bit mode on the Opteron. Even though the bitness isn't going to speed it up, 64-bit mode has more registers, which would benefit the opteron in these results. Again, with a better OS, the Opteron would be likely to take a bigger lead.
So, the benchmarks aren't perfect by any means, but they're not completely useless, either. If nothing else, it's a good way to rub Intel's nose in the fact that their overpriced chips are being beaten by bargain-priced underdogs, even when the cards are stacked in Intel's favor.
steve
I highly doubt that HP is making their own chipsets, or coming up with any other improvements that would give them a significant *hardware* advantage over the others. And with the Opteron being so self-contained, it would be even harder. With the memory controller being integrated, as long as you have the HyperTransport lines wired up, then most of the performance difference will come from memory timings and the like.
Out of the quad-Opteron arena, the Tyan is (as you imply) a complete joke, its memory design is completely, retardedly crippled. However, the vast majority of quad-Opteron boards are either AMD's design or NewIsys' design, and those two companies are betting the farm, so to speak, on the Opteron. I have a hard time believing that either of them are just lazily selling unoptimized boards.
Also, if anything, "vanilla" boards tend to be *overly* optimized, at the possible expense of stability. As an example, if you take a look at the *actual* FSB on a good number of boards, you'll find it measurably higher than what the BIOS screen told you it was.
steve
Yes, the tests weren't exactly apples-to-apples - the outcomes are actually much better for AMD than the graphs would initially appear.
The graphs mean that Opterons with a "measly" 1 meg of cache are beating out Xeons that have (a) four times the cache, (b) 50% higher clock speed, and (c) a price tag that's three times higher.
Hats off to AMD. In times past (K2/K3), price was the only thing they had better than Intel. Now they've got both price and performance.
steve
I once watched an interview with a specialist in airport security. He talked about seeing a family in one of the terminals - mother, father, kids - taking a video of themselves in the terminal before a flight. No big deal - except they showed up and did it again the next day. And the next day. And the next day - for an entire week.
What were they doing? The video camera wasn't focused on the smiling, waving family - it was zoomed to the pay phones some distance behind them, recording people punching in their credit-card numbers.
steve
This really doesn't surprise me. BSD's tend to be very secure by default, when the most popular Linux distro (RedHat) seems to strive for insecure defaults, and some other distros aren't much better. The problem isn't in Linux, it's the insecure packages that the distributors bundle up and turn on by default.
Really. If Linux distributions would turn services off by default, and leave it up to the user to turn them on, I honestly believe that the succesful attacks against Linux would be cut by at least 50%. And if they'd get away from the "classically insecure" daemons, I believe you'd cut the rate much, much farther.
steve
seem to be dominated by people using cheats
While people using cheats do exist, they're not nearly as prevalent as you seem to think, you're just unable to accept that people are that good.
Take Quake 3, for example. I got tired of having my butt handed to me by people with the railgun. So, I limitted myself to *only* using the railgun. In time, I got pretty good. With more time, I got really good. With even more time, I got fan-freaking-tastic. When you're single, don't own a home, and have a 1-megabit pipe into your apartment, it's amazing how much time you can dedicate/waste on those things!
So, what did that get me? A bunch of whining cry-babies yelling "Aimbot! Cheater!" every time I killed them. After they did it enough to annoy me, I'd chase them down and kill them with the gauntlet. At that point, they'd generally just leave the game.
steve
Does this mean they'll actually MD5 the root password?
(Sarcasm-less explanation: During the RedHat installation procedure, the ability to choose to use MD5-encrpted passwords comes *after* you choose your root password, so your root password is encrypted with much weaker encryption until you change it.)
steve
and a lot of the routing logic is hardware based so they are able to scale & perform much better than the latter
That's only true if you're not doing anything but routing. Start adding in any features that are remotely useful, and you've gone from CEF/fast switching to process switching. And that means you take a 5x to 10x performance hit.
You actually have to purchase a *VERY* expensive Cisco to get one that can't be out-performed by even a relatively modest PC.
And even on the high-end, there are PC's that will completely blow away any but the very largest offerings from Cisco.
There are, of course, several real advantages to using a Cisco router instead of a PC. First and foremost, if you have the money for it, you're going to be able to find a network interface for almost any type of network you can imagine. DSL to SONET, the interfaces are there for the buying. I'd *almost* say that they would "just work", but that's not always the case. I just had to upgrade the IOS versio on one of my Ciscos (a *paid* upgrade) to get nothing but support for an additional ethernet WIC.
Another large (perhaps HUGE) advantage is the fact that as long as you want to keep paying the money for it, there will always be someone there to back you up and make sure your problem gets resolved, even if you're not capable of fixing it yourself.
So, I'm not saying that there's no use for Cisco. I'm just saying that absolute performance (and especially price-for-performance) are not real advantages for them.
I don't know why they would want to hold back Linksys development though
You can't? Let's think about it: Do you think they'd rather sell a $200 Linksys router, or a $2,000+ Cisco router?
I know, it sounds cynical. Unfortunately, I've worked in enough corporations to know that is *exactly* how decisions are made on these things.
They'll keep the Linksys line around for the low-end market, the home users that don't demand much. But for anything above that, they're never going to let Linksys compete with their Cisco line. They're not stupid.
steve
The performance increase comes from a combination of lower memory latency (built-in memory controller) and an increased number of registers. The small number of registers on x86 chips has always been one of the main gripes people have had about the architecture.
steve
You didn't really think you could just plug your line out to your car's extra AUX IN port, did you?
Umm... actually, I did think that. My Honda Element has a nice 1/8" stereo plug that says "AUX" on it, right up on the dash. : )
steve
Maybe because spending millions of dollars on a space mission entails rigorously testing every piece of equipment which could take years before the launch?
Yes, I'm sure that it would have taken so long to test a larger CCD and a longer lens....
an extra-thousand bucks will probably get you nowhere
So, an extra twenty thousand bucks. That's still a completely insignificant drop in the bucked in relation to the rest of the mission. What, half a million just to press the "launch" button, along with six or ten years of paying hundreds (or thousands?) of people's salaries? Twenty grand wouldn't even be noticed.
steve