Vehicles tuned to efficiency seem to take a big hit from peak mileage depending on driving conditions.
Try measuring more useful terms like liters per 100km.
Say that you have two cars. Both only see 80% of their normal mileage in these conditions. In optimal conditions, one gets 10 mpg, the other gets 60 mpg. The first car will only lose 2 mpg, the second will lose 12 mpg. However, in 10 miles, the first car used an extra.25 gallons. The second car used an extra.04 gallons.
it looks to me as if the car's structural components are largely sacrificial
Yes, they are.
1. Energy that can be absorbed by the structure of a car is energy that is not converted into rebound motion. This reduces the total amount of acceleration the passengers experience.
2. Increasing the distance through which the energy is absorbed increases the amount of time during which that absorption takes place. This reduces the rate of acceleration (maximum G forces) the passengers experience.
A car will achieve the highest crash ratings by keeping the passenger cage rigidly intact with the passengers securely fastened to it, while allowing the entirety of the rest of the structure to crumple.
With that said, why the hell is it so hard to make a 5 mph bumper?
Take any GSM cell phone (others will do it, but GSM is by far the worst) and hold it near a speaker. Turn it on and you'll hear it chatter packets at the cell phone tower.
You hear this because the speakers are poorly shielded and as a result they are working as a crystal radio. Hopefully avionics are designed better than this - long wire runs should use balanced signaling (to enable common mode rejection), twisted pair wires, and shielding. Inter-device signaling should use stronger signals (i.e. lower impedances) to maintain a solid signal/noise ratio. When possible, these signals should be digitized and use forward error correction. The avionics themselves should be enclosed in a case that functions as a Faraday cage up to a few GHz. And if after all of that, your test results still show interference at any frequency, fiber optics can be used instead of wires.
A lot of planes are old. All of these techniques are older (although, granted, fiber optics, digitization, and FEC have become practical only since the mid-80's).
How often do you really max out your CPU cycles these days anyway?
Anytime it executes something. Although there are a number of things such as frequency stepping and pipelining that complicate the issue to an extent, a core is either executing instructions at its operating clockspeed or it is not. Similarly, a network interface is either transmitting data or it is not. Memory is either being accessed or it's not. A disk is either idle or it is not.
If those components operated faster, they'd finish any given task faster, leaving room for more stuff in their queues. The reduced latency of those faster components would make the system feel more responsive, and if you timed a large task, you'd note that it'd be done sooner. If "sooner" means "these video frames consistently finish rendering within the desired refresh rate" or "these VoIP packets leave the transmit queue before the next VoIP packet arrives" or "that computer finished logging in before the user got tempted to go get coffee or call the helpdesk" then that can be a big deal. But either way, those components would still be fully maxed out until the task completes.
I admit this is just pedantry, but perhaps what you meant to say is "Is it really your CPU that you're waiting on most of the time?"
Or just send a contract of your own via the EHLO string when your mail server connects to his to deliver the message. It's equally unenforceable, but it might get him to see the point if he feels like pressing the issue.
There is no way to achieve an orbit purely by firing a gun from the surface, no matter how much velocity it produces.
Folks, if you call in to our delta-v telethon now and pledge at the 11km/s level, we'll give you a lunar slingshot. That's just 2km/s more, and you cannot buy this in stores. Call now, our operators are standing by.
orders have been placed, contracts signed, contractors hired, training done, policies and protocols written, multitudes of careers exist around these infernal things... you don't just step in and undo all that with a snap decision.
The first one we consider a good thing - it is a check on the judicial branch by the executive branch. Pardoning does occasionally have undesirable outcomes (Richard Nixon comes to mind - also google for Maurice Clemmons) but we believe that people's liberties are so important that we want the legal process of taking them away to be interruptable at any point in the process - by the officer who would have otherwise arrested you. By the prosecuting attorney who is preparing the case against you. By the judge or by a single juror who remains unconvinced. By an appellate judge, if they accept your case. By the governor or president, if they don't want to keep you locked up.
Outlining all of the features of the legal due process system only serves to increase the horror of extrajudicial punishments, extraordinary renditions, and assassinations. Leaders who do those things should be prosecuted as criminals. We have a due process for that too - sadly, with a polarized two-party system, it won't ever be used properly or function correctly as a deterrent. In the meantime, we'd be wise to make it a campaign issue.
If you are a chip designer for AMD, you would definitely understand Intel's CPU patents
No, unless you know they've expired or have been licensed to you. Otherwise, you wouldn't dare look - if you blindly infringe on a patent in the course of your work, and sell a product based on it, you're potentially liable to damages. If you knowingly infringe on a patent in the course of your work, and sell a product based on it, you're potentially liable for treble damages for willful infringement.
Since it's highly likely that your product will infringe on someone's patents somewhere in its design, having the people making the product read the patents is corporate recklessness. Sure, have the lawyers look at them, but keep the actual designers oblivious.
if she goes into overtime, she only gets about 10% of her overtime pay and the other 90% is taken in taxes, so it's not worth it for her to work
How do you figure? The top marginal tax rate (for income above $373650 in 2010) is 35%, not 90%. The way the tax brackets work is that higher rates kick in at higher thresholds of income, they kick in on only the income *above* that threshold. So, for example, below $34000 the rate is 15%, above that it's 25% (assuming you're filing single). If you make $36000, you're only taxed that 25% rate on the $2000 above the threshold. There isn't any scenario where you don't want to make another $100 because you'd be paying anywhere close to that $100 in taxes.
Granted, you can make a little more if you happen to be at the edge of one of the lines on the tax table and aren't pushed to the next one. They're $50 increments, so this gets you at most $13. Enjoy your movie and popcorn.
I'm pretty impressed by how quiet their demo rack is - it'd be a challenge to get a good audio recording of a conversation right next to a full rack of air-cooled 1U servers - it's frustrating using a cell phone in most server rooms, just because of the fan noise. 1U systems are the worst simply because the form factor requires a large number of tiny fans running at high speed.
Even if there's some serious impracticalities with their approach, eliminating that fan noise is a huge selling point.
Assuming 50% efficiency of the charge-discharge cycle
This is way too low. Li-Ion batteries themselves are more like 98-99% and I would expect any drive electronics to be >95% efficient (cooling them would be impossible if they were any less efficient). Oh, and I pay more like $0.10/kWh in Washington state, not your $0.30/kWh.
I'm pretty sure these types of things are exponential in nature, IE, realtime charging for enough power at 50mph takes 100HP, at 70mph it takes 180HP or something, at 75mph it takes 220HP IE its not linear.
No, it's not (aerodynamic drag, the dominant factor at highway speeds, increases with the square of speed), however those numbers are way, way off. Let's assume you have a middle-of-the-road 30%-efficient engine. It'll use about 6.8 gallons of gas per hour to produce 100HP. At 50mph, that'd be 7.3 MPG. At 70mph/180HP that'd be 5.7 MPG. At 75mph/220HP that'd be 5 MPG.
Consider this: A Chevette will (eventually) get up to about 100mph on a straight-and-level road and only has a ~60hp engine. The Volt's genset is 55kW (74hp).
ICE -> wheels is more efficient than ICE -> generator -> electric motor -> wheels, so you would need a bigger generator than the original ICE engine.
Only if you need pedal-to-the-metal power for a longer span of time than the batteries can provide. Unless your m.o. is going fast and turning left, that's probably not the case. When you're cruising down the highway you're generally only using about 10% of your engine's output, so it's not unreasonable to expect an engine one third of the size to suffice. In addition, you can use a more efficient engine type. External combustion engines such as the Stirling and Kalina cycle engines are more efficient, can burn nearly any fuel more completely than an ICE, and have fewer moving parts, but need to warm up first and can't change their power output level very quickly.
The real win for series plug-in hybrids is that energy from your power company is a fair bit cheaper than energy from your engine crankshaft - and you only have to use the latter when you deviate from your daily routine (obviously if your normal commute is long enough to completely kill the batteries, that'd be a dealbreaker for a car like this). Also, there's a lot of mechanical parts that go away in a pure series hybrid, particularly with a motor-in-wheel design - universal joints, differentials, transmission, clutch, starter, etc., and you have some limp-home capability if the engine dies (although a "jump" would probably take quite a bit longer).
Smells like a thief stole you stuff, regardless if he/she was working for TSA or not.
Airports always tell you never to leave your baggage unattended. If a TSA employee stole your stuff, it's the TSA's fault. If the TSA left your baggage unattended and another thief stole your stuff, it's the TSA's fault. This is a security issue - if a rogue employee can get away with taking stuff from your luggage, they can put whatever they want *in* your luggage too.
Are you saying software simply can't be inventive? That you can't possibly think of something in software that anyone else couldn't have thought of, even given the exact same problem set?
"Everything has been thought of before. The trick is to think of it again." -- Johann Wolfgang von Goethe
...despite the problems, there are a number of true innovators in software.
Here's the words of one of them:
"The idea that I can be presented with a problem, set out to logically
solve it with the tools at hand, and wind up with a program that could not
be legally used because someone else followed the same logical steps some
years ago and filed for a patent on it is horrifying." -- John Carmack
If you really want to change things, to truly help fix the system, UNDERSTAND IT. Really understand it. Slashdot won't help with that.
Most of the people I talk to about this have never heard of the guild system that preceded the patent system or the constitution's stated purpose for the patent system, and have never contemplated how the way we share information today may have rendered both obsolete. They blindly assume the system is good. I don't.
Put it this way... In real life, light ranges from 0 to 1. 0 being the absolute absense of light, and 1 being the brightest thing possible. (I'm not sure we know what that is yet):)
Light levels found on earth range from nearly 0 photons per second (due to blackbody radiation you will find a complete lack of photons only in an absolute-zero temperature environment) to at least 10^13 times brighter than the surface of the sun (if you find yourself too close to an H-bomb explosion).
It is physically impossible to produce a camera with that sort of dynamic range, because it is physically impossible to produce a nuke-proof camera. Even with more pedestrian light levels (such as pointing the camera directly at the sun) you very quickly heat up the image sensor. Assuming the heat doesn't destroy the sensor, the increased sensor temperature means more noise (blackbody radiation) and you also have more stray photons reflecting off the sensor and baffles (they are not a perfect black). This noise means less contrast.
We probably differ on terminology. 2^(50/1023000)-1 =.0000488758, which corresponds to about -44.54dB minimum theoretical SNR. Since the GPS protocol isn't nearly perfect, -43dB SNR is a quite reasonable real-world figure.
if something is below the noise floor, by extension this means it simply cannot be received. Including spread spectrum.
No, it doesn't mean that at all. It does mean that your error-free bitrate will be limited to less than the bandwidth (how much less depends on how much more noise than signal you have). GPS uses 1.023MHz of bandwidth (for the civilian signal - 10.23MHz for the military one) and has a bitrate of 50 bits/sec. Typical noise levels are -110dBm and typical signal levels are -130dBm.
They measured this on the Tesla Roadster. The graphs give you a good idea of what the forces are at work without worrying about engine inefficiencies at non-optimal power outputs.
Note that the sweet spot is less than 20mph. All of the thousands-of-miles-per-gallon records from purpose-built vehicles in hypermiling competitions are set at similar speeds, using a pulse-and-glide technique.
While the sweet spot for your own car is likely slightly higher, it's not that much higher, even with the A/C on full blast. If you're comparing 50mph to 90mph, you can safely assume that you're far enough away from the sweet spot on any car that drag is the major force and that a doubling in speed will take 4 times the power (although you will get there faster so you won't be running the engine as long). If your car gets, say, 40mpg at 50mph, it'll get roughly 33mpg at 60mph, 29mpg at 70mph, 25mpg at 80mph, 22mpg at 90mph, 20mpg at 100mph, etc.
Slashdot Mirror
Try measuring more useful terms like liters per 100km.
Say that you have two cars. Both only see 80% of their normal mileage in these conditions. In optimal conditions, one gets 10 mpg, the other gets 60 mpg. The first car will only lose 2 mpg, the second will lose 12 mpg. However, in 10 miles, the first car used an extra .25 gallons. The second car used an extra .04 gallons.
Yes, they are.
1. Energy that can be absorbed by the structure of a car is energy that is not converted into rebound motion. This reduces the total amount of acceleration the passengers experience.
2. Increasing the distance through which the energy is absorbed increases the amount of time during which that absorption takes place. This reduces the rate of acceleration (maximum G forces) the passengers experience.
A car will achieve the highest crash ratings by keeping the passenger cage rigidly intact with the passengers securely fastened to it, while allowing the entirety of the rest of the structure to crumple.
With that said, why the hell is it so hard to make a 5 mph bumper?
The article you linked to is wrong. In Washington State, unlawful discharge of a laser in the first degree is a felony unless you're a juvenile without a prior.
You hear this because the speakers are poorly shielded and as a result they are working as a crystal radio. Hopefully avionics are designed better than this - long wire runs should use balanced signaling (to enable common mode rejection), twisted pair wires, and shielding. Inter-device signaling should use stronger signals (i.e. lower impedances) to maintain a solid signal/noise ratio. When possible, these signals should be digitized and use forward error correction. The avionics themselves should be enclosed in a case that functions as a Faraday cage up to a few GHz. And if after all of that, your test results still show interference at any frequency, fiber optics can be used instead of wires.
A lot of planes are old. All of these techniques are older (although, granted, fiber optics, digitization, and FEC have become practical only since the mid-80's).
Waiting may expose the plaintiff to a Doctrine of Laches defense.
How about this for a solution: Get the major speed test websites to check for this problem. The ISPs won't stop hearing about it until it's fixed.
Anytime it executes something. Although there are a number of things such as frequency stepping and pipelining that complicate the issue to an extent, a core is either executing instructions at its operating clockspeed or it is not. Similarly, a network interface is either transmitting data or it is not. Memory is either being accessed or it's not. A disk is either idle or it is not.
If those components operated faster, they'd finish any given task faster, leaving room for more stuff in their queues. The reduced latency of those faster components would make the system feel more responsive, and if you timed a large task, you'd note that it'd be done sooner. If "sooner" means "these video frames consistently finish rendering within the desired refresh rate" or "these VoIP packets leave the transmit queue before the next VoIP packet arrives" or "that computer finished logging in before the user got tempted to go get coffee or call the helpdesk" then that can be a big deal. But either way, those components would still be fully maxed out until the task completes.
I admit this is just pedantry, but perhaps what you meant to say is "Is it really your CPU that you're waiting on most of the time?"
Or just send a contract of your own via the EHLO string when your mail server connects to his to deliver the message. It's equally unenforceable, but it might get him to see the point if he feels like pressing the issue.
Folks, if you call in to our delta-v telethon now and pledge at the 11km/s level, we'll give you a lunar slingshot. That's just 2km/s more, and you cannot buy this in stores. Call now, our operators are standing by.
Yes, in fact, that's exactly how you do it.
The first one we consider a good thing - it is a check on the judicial branch by the executive branch. Pardoning does occasionally have undesirable outcomes (Richard Nixon comes to mind - also google for Maurice Clemmons) but we believe that people's liberties are so important that we want the legal process of taking them away to be interruptable at any point in the process - by the officer who would have otherwise arrested you. By the prosecuting attorney who is preparing the case against you. By the judge or by a single juror who remains unconvinced. By an appellate judge, if they accept your case. By the governor or president, if they don't want to keep you locked up.
Outlining all of the features of the legal due process system only serves to increase the horror of extrajudicial punishments, extraordinary renditions, and assassinations. Leaders who do those things should be prosecuted as criminals. We have a due process for that too - sadly, with a polarized two-party system, it won't ever be used properly or function correctly as a deterrent. In the meantime, we'd be wise to make it a campaign issue.
No, unless you know they've expired or have been licensed to you. Otherwise, you wouldn't dare look - if you blindly infringe on a patent in the course of your work, and sell a product based on it, you're potentially liable to damages. If you knowingly infringe on a patent in the course of your work, and sell a product based on it, you're potentially liable for treble damages for willful infringement.
Since it's highly likely that your product will infringe on someone's patents somewhere in its design, having the people making the product read the patents is corporate recklessness. Sure, have the lawyers look at them, but keep the actual designers oblivious.
How do you figure? The top marginal tax rate (for income above $373650 in 2010) is 35%, not 90%. The way the tax brackets work is that higher rates kick in at higher thresholds of income, they kick in on only the income *above* that threshold. So, for example, below $34000 the rate is 15%, above that it's 25% (assuming you're filing single). If you make $36000, you're only taxed that 25% rate on the $2000 above the threshold. There isn't any scenario where you don't want to make another $100 because you'd be paying anywhere close to that $100 in taxes.
Granted, you can make a little more if you happen to be at the edge of one of the lines on the tax table and aren't pushed to the next one. They're $50 increments, so this gets you at most $13. Enjoy your movie and popcorn.
I'm pretty impressed by how quiet their demo rack is - it'd be a challenge to get a good audio recording of a conversation right next to a full rack of air-cooled 1U servers - it's frustrating using a cell phone in most server rooms, just because of the fan noise. 1U systems are the worst simply because the form factor requires a large number of tiny fans running at high speed.
Even if there's some serious impracticalities with their approach, eliminating that fan noise is a huge selling point.
http://slashdot.org/article.pl?sid=05/11/24/1932233
This is way too low. Li-Ion batteries themselves are more like 98-99% and I would expect any drive electronics to be >95% efficient (cooling them would be impossible if they were any less efficient). Oh, and I pay more like $0.10/kWh in Washington state, not your $0.30/kWh.
No, it's not (aerodynamic drag, the dominant factor at highway speeds, increases with the square of speed), however those numbers are way, way off. Let's assume you have a middle-of-the-road 30%-efficient engine. It'll use about 6.8 gallons of gas per hour to produce 100HP. At 50mph, that'd be 7.3 MPG. At 70mph/180HP that'd be 5.7 MPG. At 75mph/220HP that'd be 5 MPG.
Consider this: A Chevette will (eventually) get up to about 100mph on a straight-and-level road and only has a ~60hp engine. The Volt's genset is 55kW (74hp).
Only if you need pedal-to-the-metal power for a longer span of time than the batteries can provide. Unless your m.o. is going fast and turning left, that's probably not the case. When you're cruising down the highway you're generally only using about 10% of your engine's output, so it's not unreasonable to expect an engine one third of the size to suffice. In addition, you can use a more efficient engine type. External combustion engines such as the Stirling and Kalina cycle engines are more efficient, can burn nearly any fuel more completely than an ICE, and have fewer moving parts, but need to warm up first and can't change their power output level very quickly.
The real win for series plug-in hybrids is that energy from your power company is a fair bit cheaper than energy from your engine crankshaft - and you only have to use the latter when you deviate from your daily routine (obviously if your normal commute is long enough to completely kill the batteries, that'd be a dealbreaker for a car like this). Also, there's a lot of mechanical parts that go away in a pure series hybrid, particularly with a motor-in-wheel design - universal joints, differentials, transmission, clutch, starter, etc., and you have some limp-home capability if the engine dies (although a "jump" would probably take quite a bit longer).
Airports always tell you never to leave your baggage unattended. If a TSA employee stole your stuff, it's the TSA's fault. If the TSA left your baggage unattended and another thief stole your stuff, it's the TSA's fault. This is a security issue - if a rogue employee can get away with taking stuff from your luggage, they can put whatever they want *in* your luggage too.
Let's go with a more cut-and-dried revolutionary-not-evolutionary example: Why do you think airplanes have so many French-named parts?
"Everything has been thought of before. The trick is to think of it again." -- Johann Wolfgang von Goethe
Here's the words of one of them:
"The idea that I can be presented with a problem, set out to logically solve it with the tools at hand, and wind up with a program that could not be legally used because someone else followed the same logical steps some years ago and filed for a patent on it is horrifying." -- John Carmack
Most of the people I talk to about this have never heard of the guild system that preceded the patent system or the constitution's stated purpose for the patent system, and have never contemplated how the way we share information today may have rendered both obsolete. They blindly assume the system is good. I don't.
Light levels found on earth range from nearly 0 photons per second (due to blackbody radiation you will find a complete lack of photons only in an absolute-zero temperature environment) to at least 10^13 times brighter than the surface of the sun (if you find yourself too close to an H-bomb explosion).
It is physically impossible to produce a camera with that sort of dynamic range, because it is physically impossible to produce a nuke-proof camera. Even with more pedestrian light levels (such as pointing the camera directly at the sun) you very quickly heat up the image sensor. Assuming the heat doesn't destroy the sensor, the increased sensor temperature means more noise (blackbody radiation) and you also have more stray photons reflecting off the sensor and baffles (they are not a perfect black). This noise means less contrast.
We probably differ on terminology. 2^(50/1023000)-1 = .0000488758, which corresponds to about -44.54dB minimum theoretical SNR. Since the GPS protocol isn't nearly perfect, -43dB SNR is a quite reasonable real-world figure.
No, it doesn't mean that at all. It does mean that your error-free bitrate will be limited to less than the bandwidth (how much less depends on how much more noise than signal you have). GPS uses 1.023MHz of bandwidth (for the civilian signal - 10.23MHz for the military one) and has a bitrate of 50 bits/sec. Typical noise levels are -110dBm and typical signal levels are -130dBm.
They measured this on the Tesla Roadster. The graphs give you a good idea of what the forces are at work without worrying about engine inefficiencies at non-optimal power outputs.
Note that the sweet spot is less than 20mph. All of the thousands-of-miles-per-gallon records from purpose-built vehicles in hypermiling competitions are set at similar speeds, using a pulse-and-glide technique.
While the sweet spot for your own car is likely slightly higher, it's not that much higher, even with the A/C on full blast. If you're comparing 50mph to 90mph, you can safely assume that you're far enough away from the sweet spot on any car that drag is the major force and that a doubling in speed will take 4 times the power (although you will get there faster so you won't be running the engine as long). If your car gets, say, 40mpg at 50mph, it'll get roughly 33mpg at 60mph, 29mpg at 70mph, 25mpg at 80mph, 22mpg at 90mph, 20mpg at 100mph, etc.