No, it was very intelligent to put transparent networking in X. In a working environment, this is a key feature. I frequently have to run programs which take more memory than my deskstop workstation is equipped with (256 MB). On those occasions, all I have to do is log in to one of our department servers (2+ GB) and run the application, and it pops up on my monitor just like it was local.
If it were done like a PC, where windowed apps write directly to the hardware, I would have to either log in to the machine (which would require it to be accessible, instead of residing in a server closet), or run some hack like PCs do to perform remote administration.
It made sense when X was designed, and they pictured central servers running apps for thin-client workstations. It still makes sense in large networked environments with thousands of workstations (i.e. any major corporation). It doesn't make sense for running Quake on your standalone Linux box. It would be kind of difficult to predict 20 years ago that someone would even try to do that...
Re:Boy you people like living dangerously
on
A New DeCSS
·
· Score: 1
Actually, you CAN sue the entire world. All you have to do is get all of their names and list them as defendants.
Okay, so it's a slight exaggeration, because you'd get thrown out of court. But all you have to do to sue someone in addition to your primary target is have some arguably valid reason that they are also injuring you, and be able to identify that person.
At that point, it's the defendent's responsibility to hire a lawyer and show up to defend themselves.
Sure, here you go. 1 IP address. What, you want more than 1 IP address? Well, it's going to cost you.
Go ahead, run sendmail. Oh, by the way, we blocked port 25, so it probably won't work so well. But you can use our POP servers for a nominal fee.
Your underlying assumption here is that with your IP address(s), you get unrestricted use of your connection bandwith. If that lack of restriction isn't regulated, then your connection provider can block out all services which they provide alternatives for, or provide for additional fees.
Competition? Yeah, right. I live in the middle of silicon valley, and even here many houses are too far away from their central office to get DSL. Additional broadband services are popping up, but no company is going to limit themselves to being just a bandwidth provider unless they're forced to. There's too much money in "optional" services.
I saw an article 3-4 years back about somebody trying to make a heat pipe system aimed at laptops. He had some liquid called "Flourinert" or something like that, which boiled at around 30-35 degrees C. It sounded really cool, but never heard anything after that. Then again, I was talking with an engineer at ARM, and he said the original Acorn computers came without fans (ARM processors typically consume 1 W of power), but customers were concerned that the machine would overheat, so later versions came equipped with a fan...:)
Not necessarily. It's not enough to be small to be the king of lithography -- you also have to be fast. X-ray lithography can already build insanely small circuits, but the problem is that it's a single beam. Imagine how long it would take to draw 5 million transistors on 10 layers with a pencil.
For all its limitations, photolithography can do an entire layer on multiple chips all in one shot. The more transistors you add, the faster it gets relative to a beam-method of drawing.
This plotter, although it looks very cool, also looks like its still drawing lines, as opposed to exposing portions of a wafer.
People have been predicting the death of visible-light lithography for years now, but it's so useful that people keep coming up with ways to extend its life -- I think the latest enhancement was the use of phase-shifted interference patterns to draw features with visible light that were smaller than the wavelength of the light used...
A newborn is a life. So is a plant. A fair comparison? May not sound like it, but a man I worked for had a "severly disabled son". His son was about 17 years old and had the IQ of, oh, probably around a 6-month old. In the months I worked for him, I never heard a sound out of his son's mouth that sounded anything like speech.
So to re-frame the discussion, people DO think about euthanizing their 31 old after a car accident. What do you think "turn the machines off" means?
It means that the person you knew is DEAD, because their brain is gone. It's the same discussion that Singer is trying to bring up about severly disabled babies. If there's no brain there, your child's future is nothing more than being a big drooling idiot.
And an expensive one, at that. My former employer had to remodel his bathroom so they could lower their son into the tub with a hoist, and had a custom van so they could load him in in his wheelchair. He was well-off enough be able to afford these without it hurting him too much, but having a son or daughter like this could ruin your life, and give you nothing in return. And something like that is worthy of discussion.
Okay, this really isn't that difficult a concept. I essentially use an NC every day. It isn't called an NC, of course, it's a Sun workstation. But the only thing local on it is its OS (and most of the config files for it are updated remotely), and some MP3 files I secreted away on the scratch disk.
As for "tyranny of the admins", at this company the NT machines are also under remote administration, and the users bitch constantly about it, since it does all sorts of strange things to their computers. Oh, and they STILL can't just log in to someone else's machine.
OTOH, with everyone running around saying how cheap NCs are going to be, I say: get a clue. The only difference between a PC and an NC is the possible lack of hard disk, and hard disks are only about $100-$150. Also, if you're running Linux, you'll probably need the local disk for swap. All the expensive parts of the computer (CPU, memory) are still there.
Rambus memory has high initial latency, but then has very low latency to accesses within the same line. Also, it's a burst-memory protocol, which means that although it's only 16-bits wide, you can't access less than "x" bytes (last time I looked it was 8-bits wide and always sent at least 8 words or so). It's not good for random accesses, but it's very good at sequential accesses.
What this means is that your performance depends on your data locality -- if your accesses are close to each other, you'll get much better performance.
Also remember that processors don't directly access memory; they request data from the cache, which forces a cache miss and fill. If the system is designed well, the cache line size should be the same as the minimum data size for Rambus.
His math is accurate for a large-working set scenario. Memory bandwidth is the limiting factor for any non-trivial vector computing problem. If you look at the original Cray computers, the revolutionary part of the computer wasn't just the CPU, but also the very wide, interleaved memories that allowed it to fetch new data on every clock.
The key is what you consider a "real world application". If it's Quake III, then quite possibly the full set of floating point vectors would fit in a 1-2MB L2 cache. If it's scientific computing, then it drops back to memory bus rate.
Slashdot Mirror
No, it was very intelligent to put transparent networking in X. In a working environment, this is a key feature. I frequently have to run programs which take more memory than my deskstop workstation is equipped with (256 MB). On those occasions, all I have to do is log in to one of our department servers (2+ GB) and run the application, and it pops up on my monitor just like it was local.
If it were done like a PC, where windowed apps write directly to the hardware, I would have to either log in to the machine (which would require it to be accessible, instead of residing in a server closet), or run some hack like PCs do to perform remote administration.
It made sense when X was designed, and they pictured central servers running apps for thin-client workstations. It still makes sense in large networked environments with thousands of workstations (i.e. any major corporation). It doesn't make sense for running Quake on your standalone Linux box. It would be kind of difficult to predict 20 years ago that someone would even try to do that...
Actually, you CAN sue the entire world. All you have to do is get all of their names and list them as defendants.
Okay, so it's a slight exaggeration, because you'd get thrown out of court. But all you have to do to sue someone in addition to your primary target is have some arguably valid reason that they are also injuring you, and be able to identify that person.
At that point, it's the defendent's responsibility to hire a lawyer and show up to defend themselves.
Sure, here you go. 1 IP address. What, you want more than
1 IP address? Well, it's going to cost you.
Go ahead, run sendmail. Oh, by the way, we blocked port
25, so it probably won't work so well. But you can use our
POP servers for a nominal fee.
Your underlying assumption here is that with your IP
address(s), you get unrestricted use of your connection
bandwith. If that lack of restriction isn't regulated, then
your connection provider can block out all services which
they provide alternatives for, or provide for additional
fees.
Competition? Yeah, right. I live in the middle of silicon
valley, and even here many houses are too far away from
their central office to get DSL. Additional broadband
services are popping up, but no company is going to
limit themselves to being just a bandwidth provider unless
they're forced to. There's too much money in "optional"
services.
I saw an article 3-4 years back about somebody trying to make a heat pipe system aimed at laptops. He had some liquid called "Flourinert" or something like that, which boiled at around 30-35 degrees C. It sounded really cool, but never heard anything after that. Then again, I was talking with an engineer at ARM, and he said the original Acorn computers came without fans (ARM processors typically consume 1 W of power), but customers were concerned that the machine would overheat, so later versions came equipped with a fan... :)
Not necessarily. It's not enough to be small to be the king of lithography -- you also have to be fast. X-ray lithography can already build insanely small circuits, but the problem is that it's a single beam. Imagine how long it would take to draw 5 million transistors on 10 layers with a pencil.
For all its limitations, photolithography can do an entire layer on multiple chips all in one shot. The more transistors you add, the faster it gets relative to a beam-method of drawing.
This plotter, although it looks very cool, also looks like its still drawing lines, as opposed to exposing portions of a wafer.
People have been predicting the death of visible-light lithography for years now, but it's so useful that people keep coming up with ways to extend its life -- I think the latest enhancement was the use of phase-shifted interference patterns to draw features with visible light that were smaller than the wavelength of the light used...
A newborn is a life. So is a plant. A fair comparison? May not sound like it, but a man I worked for had a "severly disabled son". His son was about 17 years old and had the IQ of, oh, probably around a 6-month old. In the months I worked for him, I never heard a sound out of his son's mouth that sounded anything like speech.
So to re-frame the discussion, people DO think about euthanizing their 31 old after a car accident. What do you think "turn the machines off" means?
It means that the person you knew is DEAD, because their brain is gone. It's the same discussion that Singer is trying to bring up about severly disabled babies. If there's no brain there, your child's future is nothing more than being a big drooling idiot.
And an expensive one, at that. My former employer had to remodel his bathroom so they could lower their son into the tub with a hoist, and had a custom van so they could load him in in his wheelchair. He was well-off enough be able to afford these without it hurting him too much, but having a son or daughter like this could ruin your life, and give you nothing in return. And something like that is worthy of discussion.
Okay, this really isn't that difficult a concept. I essentially use an NC every day. It isn't called an NC, of course, it's a Sun workstation. But the only thing local on it is its OS (and most of the config files for it are updated remotely), and some MP3 files I secreted away on the scratch disk.
As for "tyranny of the admins", at this company the NT machines are also under remote administration, and the users bitch constantly about it, since it does all sorts of strange things to their computers. Oh, and they STILL can't just log in to someone else's machine.
OTOH, with everyone running around saying how cheap NCs are going to be, I say: get a clue. The only difference between a PC and an NC is the possible lack of hard disk, and hard disks are only about $100-$150. Also, if you're running Linux, you'll probably need the local disk for swap. All the expensive parts of the computer (CPU, memory) are still there.
Rambus memory has high initial latency, but then has very low latency to accesses within the same line. Also, it's a burst-memory protocol, which means that although it's only 16-bits wide, you can't access less than "x" bytes (last time I looked it was 8-bits wide and always sent at least 8 words or so). It's not good for random accesses, but it's very good at sequential accesses.
What this means is that your performance depends on your data locality -- if your accesses are close to each other, you'll get much better performance.
Also remember that processors don't directly access memory; they request data from the cache, which forces a cache miss and fill. If the system is designed well, the cache line size should be the same as the minimum data size for Rambus.
His math is accurate for a large-working set scenario. Memory bandwidth is the limiting factor for any non-trivial vector computing problem. If you look at the original Cray computers, the revolutionary part of the computer wasn't just the CPU, but also the very wide, interleaved memories that allowed it to fetch new data on every clock.
The key is what you consider a "real world application". If it's Quake III, then quite possibly the full set of floating point vectors would fit in a 1-2MB L2 cache. If it's scientific computing, then it drops back to memory bus rate.