This isn't true. The limit is based on your Effective Isotropic Radiated Power, or EIRP. This is formulated by the combination of your output amplifier and your antenna. To get your EIRP, simply add the output power of your amplifier, expressed in dBm (dB referred to 1 milliwatt), to the gain of your antenna, expressed in dB.
So, if you have a transmitter with an output power of +14dBm, and an antenna with a gain of 18dB, then you have an EIRP of 14+18 = 32dBm, which is almost 2 watts.
If you have a transmitter with an output power of 14dBm, an amplifier with a gain of +10dB, and an antenna with a gain of +15dB, then you end up with an EIRP of 14+10+15 = 39dBm. So, in extremely simplified terms, you simply add up all the dB's to get your EIRP.
To convert your EIRP into a "wattage" number, you divide your dBm by 10, and then raise 10 to that number. So, if you have 36dBm, you would do 10**3.6, which is 3.981 Watts.
Some useful things to remember is that adding 10dB is the same as multiplying the output power by a factor of 10. Adding 3dB is the same as doubling your output power. Likewise, subtracting 3dB is halving and subtracting 10dB is decimating.
There are two limits, one for point to multipoint and the other for point to point links. For the former, you're allowed up to 36dBm EIRP, which is to say, 36dB over 1 milliwatt, which is 4 Watts EIRP. For the latter, you're allowed 48dBm, which is just a tiny bit over 50 Watts (50.118).
Be careful what antennae you use if you're in the US. For 802.11 in America, point to multipoint links are limited to 36dBm EIRP, which would be exceeded by 8dB using a 24dBi antenna with a WAP11. However, you're allowed 48dBm EIRP for a point to point link.
Also, before you go run out and buy a high gain antenna, be mindful that any consumer access point has TWO antennae - one for transmit and one for receive, so you'll need two high gain antennae. Access points that only have one antenna actually do have two. The receiving antenna is external and the transmitting antenna is an internal stubby. here is a picture of the inside of one of my D-Link 900AP+ access points. You can see the stubby transmit antenna on the right.
What I did for my long-haul link is remove the coax and connector from another access point, and swapped it with the stubby. This leaves me with one access point having two external antenna jacks and one with two internal stubbies. The latter is fine for use around the house while the former handles my local hotspot. Here is a picture of my antenna pair.
I think I was able to put together a single AP/Antenna pair for about $700, which included the two antennae, the custom cables, the two AP's required to make one with two jacks, and all the necessary hardware, weatherproof enclosures, mounting frame, power over ethernet (homemade), and support structure.
The longest link I've been able to establish to date is a 9.1mile link at 1Mbit, which survived for about 45 seconds. I can reliably make 2 miles at 22Mbit as long as the other side has the same setup.
Part 15 of the FCC rules, with which all 802.11 equipment must comply, state in part that "This device must accept any interference, including that which may cause undesired operation."
Federal law preempts the ability to complain about interference in the ISM band. This is unlicensed spectrum to which no one individual or entity has any elevated use rights over any other individual or entity. This is good for adoption, although overcrowding may become an issue.
Even if only consequential, Part 15 protects joe blow average user from any recourse from a big corporation that wants to use the spectrum.
I will probably have first-hand experience with this shortly, as my new free public hotspot is aimed directly at the Starbucks down the street from my eighth floor condo.:)
Unfortunately for me, the law only allows 36dBm EIRP for point to multipoint links, so I'm trying to get a point-to-point link with a friend of mine who lives just a little further down the street, in which case I'm allowed 48dBm EIRP (which for those of you who don't know logarithmic math, is almost 16 times as much power):)
And yes, I do have a businees connection of which I am perfectly allowed to resell the bandwidth... Here's a picture of my antenna array, just before I added the isolation fence and other goodies to prevent coupling..
Would you want to spend the kind of money that you would have to to get a 1 milliohm wire resistance, just to move your PC power supply somewhere else?
It certainly wouldn't be worth it to me! I have a 3 foot run of #10AWG from my UPS to a bank of Gel Cells... even that piece of wire rises 10C with a full load of 40 amps.
Switching power supplies are generally designed with only enough ESR in the output caps to make the thing stable. Depending on the switch frequency, there can be lots of ripple current flowing through them. A general rule is that there will be 0.4 times the output current flowing through the output caps as ripple current. If you have a 10-amp supply, count on 4 amps of ripple current. Now, if you're thinking of your ohms law, you wonder how much power is dissipated because of that. The answer is: none that is due to the capacitance. All the capacitor does is change the phase of the power relative to the voltage (power factor). However, capacitors have an ESR (equivalent series resistance) that will dissipate some real power from the ripple current (and put a ripple voltage on the output). Typical ESR numbers for caps used in switchmode supplies range from 5 to 15 to 50 milliohm. So your 4 amp supply could be dissipating enough power to heat up your caps...
Caps are constantly a problem in switching supplies, because a bad one CAN explode violently, especially if it's Tantalum or Aluminum Electrolytic. There are some new Niobium oxides that have a much lower ignition energy than Tantalum, but their ESR characteristics are somewhat less ideal for switchmode power supply applications.
Another thing to note about switchmode power supplies none of the caps should have a high enough voltage to hurt anybody. Usually, the input transformer steps down the line voltage to a level suitable for input capacitors, generally something at least 20% less than 35V (the popular rating for input caps).
Let's say you have your system happily sitting there with all of its stuff humming along. Chances are, it's using about 200W of power. Let's say for sake of argument that it's using an amp at 12V (12W), 5 amps at 5V (25W), and the remaining 162.5W at 3.3V. That 3.3V rail is supplying just shy of 50 amps. 50 amps is a whole hell of a lot to be pushing through a long wire. Even if your wire only had 0.01ohm of resistance, you'd be dropping half a volt and end up with only 2.8V at the computer. You'd also be dissipating 25W of energy in the wire itself as heat.
Here's are some interesting experiments.
Experiment 1: Use a thermocouple to measure the steady-state ambient temperature of the inside of your computer. Then, turn your computer off, let it cool to room temp, and attach a thermocouple to the 3.3V wire coming out of your power supply. Then, turn it on and start up your favorite CPU-pegging application. After a while, note how much hotter than the internal ambient air the 3.3V wire is. It's likely to be at least 5-10C higher.
Experiment 2: Unplug the power supply from everything in the computer. Turn the power on and measure the zero-load (open circuit) voltage of the 3.3V rail (it's probably something more like 3.45V). Then, plug everything back in, start the computer, and start up your second-favorite CPU-pegging application. Then, use your multimeter to measure the 3.3V wire at the motherboard. You should be able to get to it on the connector where the wire is attached. It's probably a lot closer to 3.3V.
Just a couple of things to demonstrate voltage drop and power loss in high current applications...
Mechanically more elegant, but not electrically sound. The addition of the wires connecting the power supply switch FETs to the board was the wrong thing to do. Switching power supplies use closed-loop feedback to maintain the proper voltage on the output. Adding long wires such as those can add enough inductance to either cause the poles to move around such that there is the incorrect amount of phase margin in the feedback loop, or to add additional poles which the loop compensation circuitry is not designed to handle. This can cause the output of the switching supply to oscillate, or at the very least, exhibit a significantly higher ripple amplitude. He added at least 12" of wire in there, which is hopefully a low enough gauge to minimize this effect, but in general, DO NOT DO THAT.
Of course, how much does it actually COST microsoft to stamp out $1.1B "worth" of software? A full copy of office is what, $750? It costs them what, $5 to package it?
Microsoft gets to write of $1.1B in losses, but it only costs them about $1.5MILLION to do it. That's going to be great for shareholder value!
Plus, they'll probably book $50 per copy for support at one point or another.. So this is just another great money-making scheme that is good for M$ and BAD for consumers.
What Microsoft should be forced to do is buy each and every californian a nice retail packaged copy of RedHat with OpenOffice, along with 4 hours of RedHat phone support and a decent linux book for newbies (if there exists such a thing).
That would teach M$ a lesson. This settlement will actually turn out to be very profitable for them.
Since you can have a different page open in each bank... should allow for some graphics intensive computation, and good multitasking. The fact that rambus can have different pages open in each back lends well to SMP, too, so maybe they're thinking of using more than one CPU?
But, anyone who actually believes Microsoft actually wants to be their "partner" in bringing "new products" to "market" is a blithering IDIOT. Microsoft isn't interested in being anyone's partner. M$ has enough money to go out and start its own mobile phone company. It's just cheaper and easier to spend $12M to steal the research and IP.
These little startups, in their eagerness to play "big company" to impress their fourbucks-going friends, will ink any deal that brings in money, because that's all they see. They don't think ahead, and don't have any idea whom their friends and enemies are. Microsoft was probably interting and rotating the knife in their backs before the ink was even dry...
I've run the "Hotmail Test" several times and every time, I get spam within 4-5 days of opening the account. Even if I never ever send an email, the amount of spam grows approximately linearly with time... it only takes about 2 months to exhaust your 2MB quota daily....
At least that was the case the last time I ran this little experiment...
It's no secret, at least it shouldn't be, that Micro$oft is making money selling your hotmail address (yet then they spam you with advertisements for their spam-blocking software)...
This sounds like a nice idea, but it's not the same as developing a _well-optimized_ compiler. Don't get me wrong, if I were legally able to do so, I would love to contribute to something like this (however, I can't because I can't (and won't) assign the work to them - it ain't "free" if someone owns it). But, writing a compiler for a different language won't make it optimize F77 any better.:)
I do plan on testing their binaries, however, and seeing if I can at least help that way..
This guy _really_ loves his wife. All criticisms of his technical prowess aside, this was a completely amazing project that should be commended. I am in awe as much as I am jealous of all of the brownie points this guy must have been given on Christmas morning...
Basically, they have a _very_ loose definition of a multiprocessor machine, and binaries cannot be run on computers or computer clusters containing more than 4 CPUs aggregate. Unfortunatley, their definition of a multiprocessor machine is so vague that it would include a distributed computing platform, such as the type that is used at distributed.net.
Since one of my main areas of research is distributed computing for the solution of electromagnetic problems, this precludes me from using their compiler. Imagine if I started something like d.net for electromagnetics, and 4 million people signed up, by their licensing terms I would have to buy 4 million licenses at $249 a pop.
I know what you're going to say: "Oh that can't possibly be what they meant," and I'll concede that you're probably right, but according to my attorney, that's what the document says, so that's that.
Here is a more in-depth comparison of Fortran 90 compilers for linux. They compared Intel, NAG, Lahey, and a couple of other compilers. Here is a comparison of Fortran 77 compilers from the same folks. GNU g77 is actually the slowest of them all, and I've actually confirmed that it is the slowest of a group consisting of DEC/Win32, Lahey/Linux, and g77. I've always dreamed of the day that open source developers would throw some real brainweight at a really well optimized Fortran compiler for linux, but it looks like I'll just have to keep dreaming. Lahey is only $199 or so, but they place some HORRIBLE licensing restrictions on the binaries that are created with their compiler. The DEC/Win32 compiler is also really nice, but since I'm not in school anymore, I'm not licensed to use it, and even if I _wanted_ to whore myself out to Micro$oft, I couldn't afford to.
Just to put some things into perspective, here are some numerical results. These were obtained on my dual-athlon 1.4GHz w/ 1GB of RAM. The task was to compute the TE and TM surface currents induced on a circular cylinder 10 wavelengths in circumferece and having a relative permittivity equal to 4-j2. The program simultaneously solves the perfect electric conducting case. The surface was discretized into 60 cells using 120 unknowns (way overkill, but just to prove the point) using the Integral Equation Asymptotic Phase method.
As you can see, there's really no comparison, except that the lahey-created binary uses about 10% more RAM than does the one created with g77. This is just a summary comparison as I did not go into measuring the difference in the error of the two results compared to a reference solution. I'm assuming that both solutions are about the same with regard to accuracy.
What is really happening here
on
DOD vs. 802.11b
·
· Score: 1
AOL, MSN, Comcast: "Hmmm, we don't like 802.11, wireless bad, cable/dsl good, free internet bad, profits good"
AOL, MSN, Comcast: "Hmmm, government has power. We want power. Let's buy power."
AOL, MSN, Comcast: "Hmmm, war on terror, government gets whatever it wants for war on terror, war on terror good, wireless bad."
AOL, MSN, Comcast: "Let's bombard DoD with FUD, Dod will also think wireless bad, cable/dsl good."
AOL, MSN, Comcast: "Wireless in trouble with government now. Good. Wireless bad, cable/dsl good."
It should be pointed out that this study was commissioned by MicroSoft, so you have to consider that it may have been biased. My version of the article submission pointed this out:) Yes, this thing DID get submitted a lot, I'm sure.
How am I going to goof off while pretending to work now? Now I have one fewer thing to do to distract myself from my real work:(
All joking aside, I liked electronic news a LOT more than I liked EETimes. EN actually sent me relevant, interesting stories whereas EETimes only publishes articles based on the IEEE's agenda du jour
Am I the only one who notices that EETimes articles are almost always political in nature whereas EN just reports the facts? I will miss EN for sure.
This ruling may have a profound effect on the PanIP Case. Basically, this ruling prevents a litigant from dragging a defendant to California in order to defend themselves unless the defendant does something specifically to smite a specific entity in California:
"Thus, the purposeful availment requirement ensures that a defendant will not be haled into a jurisdiction soleley as aresult of 'random', 'fortuitous', or 'attenuated' contacts, or of the 'unilateral activity of another party or person'."
To me, this means basically that if you only have random, casual contact with California, then you can't be dragged over there to defend yourself in a case.
I find it interesting that there are many places in this ruling that show a "[Citation]" where the judges obviously meant to insert citations, but forgot. I hope that doesn't open this ruling up for challenge - maybe it's just a draft...
So, the PanIP defendants can't be required to come to California to defend their cases. This should make it much harder for PanIP and cheaper for the defendants. If I were the PanIP defense lawyers, I'd be all over this like flies on... well.. you know..
Then again, IANAL, so I could be completely wrong on this..
Clerk: Hi, a couple of batteries huh? Can i have your name and address? Joe Blow: Uhh, Stevens? Clerk: Address? Joe Blow: It's probably already in the computer, try under Dick, my girlfriend's cousin.. Clerk: ahh here it is, that'll be $2.14 please Joe Blow: Here ya go, thanks
Of Course, Joe's ex-girlfriend, Jane, who just broke up with him 3 hours before because he's been beating her for the past 6 months, is hiding at her cousin's house. Joe is now walking out of the store with her cousin's street address on his receipt so he can go over and "tell her how sorry he is" with a 9mm handgun...
Slashdot Mirror
This isn't true. The limit is based on your Effective Isotropic Radiated Power, or EIRP. This is formulated by the combination of your output amplifier and your antenna. To get your EIRP, simply add the output power of your amplifier, expressed in dBm (dB referred to 1 milliwatt), to the gain of your antenna, expressed in dB.
So, if you have a transmitter with an output power of +14dBm, and an antenna with a gain of 18dB, then you have an EIRP of 14+18 = 32dBm, which is almost 2 watts.
If you have a transmitter with an output power of 14dBm, an amplifier with a gain of +10dB, and an antenna with a gain of +15dB, then you end up with an EIRP of 14+10+15 = 39dBm. So, in extremely simplified terms, you simply add up all the dB's to get your EIRP.
To convert your EIRP into a "wattage" number, you divide your dBm by 10, and then raise 10 to that number. So, if you have 36dBm, you would do 10**3.6, which is 3.981 Watts.
Some useful things to remember is that adding 10dB is the same as multiplying the output power by a factor of 10. Adding 3dB is the same as doubling your output power. Likewise, subtracting 3dB is halving and subtracting 10dB is decimating.
There are two limits, one for point to multipoint and the other for point to point links. For the former, you're allowed up to 36dBm EIRP, which is to say, 36dB over 1 milliwatt, which is 4 Watts EIRP. For the latter, you're allowed 48dBm, which is just a tiny bit over 50 Watts (50.118).
Don't you mean WLAN parties? :)
Also, before you go run out and buy a high gain antenna, be mindful that any consumer access point has TWO antennae - one for transmit and one for receive, so you'll need two high gain antennae. Access points that only have one antenna actually do have two. The receiving antenna is external and the transmitting antenna is an internal stubby. here is a picture of the inside of one of my D-Link 900AP+ access points. You can see the stubby transmit antenna on the right.
What I did for my long-haul link is remove the coax and connector from another access point, and swapped it with the stubby. This leaves me with one access point having two external antenna jacks and one with two internal stubbies. The latter is fine for use around the house while the former handles my local hotspot. Here is a picture of my antenna pair.
I think I was able to put together a single AP/Antenna pair for about $700, which included the two antennae, the custom cables, the two AP's required to make one with two jacks, and all the necessary hardware, weatherproof enclosures, mounting frame, power over ethernet (homemade), and support structure.
The longest link I've been able to establish to date is a 9.1mile link at 1Mbit, which survived for about 45 seconds. I can reliably make 2 miles at 22Mbit as long as the other side has the same setup.
Federal law preempts the ability to complain about interference in the ISM band. This is unlicensed spectrum to which no one individual or entity has any elevated use rights over any other individual or entity. This is good for adoption, although overcrowding may become an issue.
Even if only consequential, Part 15 protects joe blow average user from any recourse from a big corporation that wants to use the spectrum.
I will probably have first-hand experience with this shortly, as my new free public hotspot is aimed directly at the Starbucks down the street from my eighth floor condo. :)
Unfortunately for me, the law only allows 36dBm EIRP for point to multipoint links, so I'm trying to get a point-to-point link with a friend of mine who lives just a little further down the street, in which case I'm allowed 48dBm EIRP (which for those of you who don't know logarithmic math, is almost 16 times as much power) :)
And yes, I do have a businees connection of which I am perfectly allowed to resell the bandwidth... Here's a picture of my antenna array, just before I added the isolation fence and other goodies to prevent coupling..
Would you want to spend the kind of money that you would have to to get a 1 milliohm wire resistance, just to move your PC power supply somewhere else?
It certainly wouldn't be worth it to me! I have a 3 foot run of #10AWG from my UPS to a bank of Gel Cells... even that piece of wire rises 10C with a full load of 40 amps.
Actually...
Switching power supplies are generally designed with only enough ESR in the output caps to make the thing stable. Depending on the switch frequency, there can be lots of ripple current flowing through them. A general rule is that there will be 0.4 times the output current flowing through the output caps as ripple current. If you have a 10-amp supply, count on 4 amps of ripple current. Now, if you're thinking of your ohms law, you wonder how much power is dissipated because of that. The answer is: none that is due to the capacitance. All the capacitor does is change the phase of the power relative to the voltage (power factor). However, capacitors have an ESR (equivalent series resistance) that will dissipate some real power from the ripple current (and put a ripple voltage on the output). Typical ESR numbers for caps used in switchmode supplies range from 5 to 15 to 50 milliohm. So your 4 amp supply could be dissipating enough power to heat up your caps...
Caps are constantly a problem in switching supplies, because a bad one CAN explode violently, especially if it's Tantalum or Aluminum Electrolytic. There are some new Niobium oxides that have a much lower ignition energy than Tantalum, but their ESR characteristics are somewhat less ideal for switchmode power supply applications.
Another thing to note about switchmode power supplies none of the caps should have a high enough voltage to hurt anybody. Usually, the input transformer steps down the line voltage to a level suitable for input capacitors, generally something at least 20% less than 35V (the popular rating for input caps).
Here's why it won't work:
Let's say you have your system happily sitting there with all of its stuff humming along. Chances are, it's using about 200W of power. Let's say for sake of argument that it's using an amp at 12V (12W), 5 amps at 5V (25W), and the remaining 162.5W at 3.3V. That 3.3V rail is supplying just shy of 50 amps. 50 amps is a whole hell of a lot to be pushing through a long wire. Even if your wire only had 0.01ohm of resistance, you'd be dropping half a volt and end up with only 2.8V at the computer. You'd also be dissipating 25W of energy in the wire itself as heat.
Here's are some interesting experiments.
Experiment 1: Use a thermocouple to measure the steady-state ambient temperature of the inside of your computer. Then, turn your computer off, let it cool to room temp, and attach a thermocouple to the 3.3V wire coming out of your power supply. Then, turn it on and start up your favorite CPU-pegging application. After a while, note how much hotter than the internal ambient air the 3.3V wire is. It's likely to be at least 5-10C higher.
Experiment 2: Unplug the power supply from everything in the computer. Turn the power on and measure the zero-load (open circuit) voltage of the 3.3V rail (it's probably something more like 3.45V). Then, plug everything back in, start the computer, and start up your second-favorite CPU-pegging application. Then, use your multimeter to measure the 3.3V wire at the motherboard. You should be able to get to it on the connector where the wire is attached. It's probably a lot closer to 3.3V.
Just a couple of things to demonstrate voltage drop and power loss in high current applications...
Mechanically more elegant, but not electrically sound. The addition of the wires connecting the power supply switch FETs to the board was the wrong thing to do. Switching power supplies use closed-loop feedback to maintain the proper voltage on the output. Adding long wires such as those can add enough inductance to either cause the poles to move around such that there is the incorrect amount of phase margin in the feedback loop, or to add additional poles which the loop compensation circuitry is not designed to handle. This can cause the output of the switching supply to oscillate, or at the very least, exhibit a significantly higher ripple amplitude. He added at least 12" of wire in there, which is hopefully a low enough gauge to minimize this effect, but in general, DO NOT DO THAT.
:)
Otherwise, it looks beautiful
Of course, how much does it actually COST microsoft to stamp out $1.1B "worth" of software? A full copy of office is what, $750? It costs them what, $5 to package it?
Microsoft gets to write of $1.1B in losses, but it only costs them about $1.5MILLION to do it. That's going to be great for shareholder value!
Plus, they'll probably book $50 per copy for support at one point or another.. So this is just another great money-making scheme that is good for M$ and BAD for consumers.
What Microsoft should be forced to do is buy each and every californian a nice retail packaged copy of RedHat with OpenOffice, along with 4 hours of RedHat phone support and a decent linux book for newbies (if there exists such a thing).
That would teach M$ a lesson. This settlement will actually turn out to be very profitable for them.
Since you can have a different page open in each bank... should allow for some graphics intensive computation, and good multitasking. The fact that rambus can have different pages open in each back lends well to SMP, too, so maybe they're thinking of using more than one CPU?
Food for thought..
But, anyone who actually believes Microsoft actually wants to be their "partner" in bringing "new products" to "market" is a blithering IDIOT. Microsoft isn't interested in being anyone's partner. M$ has enough money to go out and start its own mobile phone company. It's just cheaper and easier to spend $12M to steal the research and IP.
These little startups, in their eagerness to play "big company" to impress their fourbucks-going friends, will ink any deal that brings in money, because that's all they see. They don't think ahead, and don't have any idea whom their friends and enemies are. Microsoft was probably interting and rotating the knife in their backs before the ink was even dry...
Because it was what was cheap enough to stack 1000 tons of on a big pile and detonate to make said yardstick...
Anyone ever see that documentary? It's been a while for me - but if anyone has a link, I'd love to see it/order it again....
I've run the "Hotmail Test" several times and every time, I get spam within 4-5 days of opening the account. Even if I never ever send an email, the amount of spam grows approximately linearly with time... it only takes about 2 months to exhaust your 2MB quota daily....
At least that was the case the last time I ran this little experiment...
It's no secret, at least it shouldn't be, that Micro$oft is making money selling your hotmail address (yet then they spam you with advertisements for their spam-blocking software)...
*sigh*
This sounds like a nice idea, but it's not the same as developing a _well-optimized_ compiler. Don't get me wrong, if I were legally able to do so, I would love to contribute to something like this (however, I can't because I can't (and won't) assign the work to them - it ain't "free" if someone owns it). But, writing a compiler for a different language won't make it optimize F77 any better. :)
I do plan on testing their binaries, however, and seeing if I can at least help that way..
This guy _really_ loves his wife. All criticisms of his technical prowess aside, this was a completely amazing project that should be commended. I am in awe as much as I am jealous of all of the brownie points this guy must have been given on Christmas morning...
Basically, they have a _very_ loose definition of a multiprocessor machine, and binaries cannot be run on computers or computer clusters containing more than 4 CPUs aggregate. Unfortunatley, their definition of a multiprocessor machine is so vague that it would include a distributed computing platform, such as the type that is used at distributed.net.
Since one of my main areas of research is distributed computing for the solution of electromagnetic problems, this precludes me from using their compiler. Imagine if I started something like d.net for electromagnetics, and 4 million people signed up, by their licensing terms I would have to buy 4 million licenses at $249 a pop.
I know what you're going to say: "Oh that can't possibly be what they meant," and I'll concede that you're probably right, but according to my attorney, that's what the document says, so that's that.
-ffast-math gives up IEEE compliance and some accuracy for speed.... which is probably why most folks don't use it for real applications..
Just to put some things into perspective, here are some numerical results. These were obtained on my dual-athlon 1.4GHz w/ 1GB of RAM. The task was to compute the TE and TM surface currents induced on a circular cylinder 10 wavelengths in circumferece and having a relative permittivity equal to 4-j2. The program simultaneously solves the perfect electric conducting case. The surface was discretized into 60 cells using 120 unknowns (way overkill, but just to prove the point) using the Integral Equation Asymptotic Phase method.
g77 Compiler (-O2 -malign-double -funroll-loops): 24.11s
Lahey Compiler (equivalent paramters): 16.45s
As you can see, there's really no comparison, except that the lahey-created binary uses about 10% more RAM than does the one created with g77. This is just a summary comparison as I did not go into measuring the difference in the error of the two results compared to a reference solution. I'm assuming that both solutions are about the same with regard to accuracy.
AOL, MSN, Comcast: "Hmmm, we don't like 802.11, wireless bad, cable/dsl good, free internet bad, profits good"
AOL, MSN, Comcast: "Hmmm, government has power. We want power. Let's buy power."
AOL, MSN, Comcast: "Hmmm, war on terror, government gets whatever it wants for war on terror, war on terror good, wireless bad."
AOL, MSN, Comcast: "Let's bombard DoD with FUD, Dod will also think wireless bad, cable/dsl good."
AOL, MSN, Comcast: "Wireless in trouble with government now. Good. Wireless bad, cable/dsl good."
It should be pointed out that this study was commissioned by MicroSoft, so you have to consider that it may have been biased. My version of the article submission pointed this out :) Yes, this thing DID get submitted a lot, I'm sure.
The probability of failure goes like this:
The probabilility of both trains failing is:
P(1st train fails) * P(2nd train fails) = 0.01
The probability of neither train failing is:
P(1st train doesn't fail) * P(2nd train...) = 0.81
The probability of exactly one train failing is:
P(1st train fails) * P(2nd train doesn't)
+ P(1st train doesnt) * P(2nd train does)
= 2 * (0.1 * 0.9) = 0.18
(notice this adds up to 1, so far)
and the probaility of at LEAST one train failiing is P(exactly one fails) + P(both fail) = 0.19
QED
All joking aside, I liked electronic news a LOT more than I liked EETimes. EN actually sent me relevant, interesting stories whereas EETimes only publishes articles based on the IEEE's agenda du jour
Am I the only one who notices that EETimes articles are almost always political in nature whereas EN just reports the facts? I will miss EN for sure.
This ruling may have a profound effect on the PanIP Case. Basically, this ruling prevents a litigant from dragging a defendant to California in order to defend themselves unless the defendant does something specifically to smite a specific entity in California:
"Thus, the purposeful availment requirement ensures that a defendant will not be haled into a jurisdiction soleley as aresult of 'random', 'fortuitous', or 'attenuated' contacts, or of the 'unilateral activity of another party or person'."
To me, this means basically that if you only have random, casual contact with California, then you can't be dragged over there to defend yourself in a case.
I find it interesting that there are many places in this ruling that show a "[Citation]" where the judges obviously meant to insert citations, but forgot. I hope that doesn't open this ruling up for challenge - maybe it's just a draft...
So, the PanIP defendants can't be required to come to California to defend their cases. This should make it much harder for PanIP and cheaper for the defendants. If I were the PanIP defense lawyers, I'd be all over this like flies on... well.. you know..
Then again, IANAL, so I could be completely wrong on this..
Clerk: Hi, a couple of batteries huh? Can i have your name and address?
Joe Blow: Uhh, Stevens?
Clerk: Address?
Joe Blow: It's probably already in the computer, try under Dick, my girlfriend's cousin..
Clerk: ahh here it is, that'll be $2.14 please
Joe Blow: Here ya go, thanks
Of Course, Joe's ex-girlfriend, Jane, who just broke up with him 3 hours before because he's been beating her for the past 6 months, is hiding at her cousin's house. Joe is now walking out of the store with her cousin's street address on his receipt so he can go over and "tell her how sorry he is" with a 9mm handgun...
The correct reference is "I guess he's not going to a minimum security, white collar resort. He's going to federal pound-me-in-the-arse prison" :)