Almost certainly completely hidden by the driver interface. Firstly because it would be easy to do, secondly because I doubt that (e.g.) the SATA interface has commands to handle this new technology invented about eight years after it was designed, and thirdly because keeping it built-in vastly simplifies the power fail handling.
This is exactly how I thought that evolution was supposed to work. Environment changes due to new predator, species evolves to handle changed environment. Yet another routine confirmation of Darwin.
How are longer legs "behaviour"? Not that I would be surprised that behaviour has Darwinian consequences. The behaviour of reproductive mating has considerable consequences - species that stop reproducing lose the evolutionary challenge.
So the Chinese and Taiwanese are dominating low-margin consumer goods, which you come across every day.
Now look where your high-margin, high tech gear is coming from. Who makes your aircraft, routers, heavy engineering equipment? The sort of thing you only see in a professional environment, and which has twenty times the markup of consumer goods. Overwhelmingly from the US, with Japan and the EU following behind.
China is doing unskilled and low skilled work. The "West" is still holding in there well on the lucrative tech stuff, and will continue to do so for some years yet. But don't rest on your laurels - if you stop running, others will overtake.
JWST [wikipedia.org]. However it only does infrared imaging, whereas Hubble covers the visible spectrum.
"Only" does infrared? Actually, there is more information to be gathered in the infrared than there is in the visible. Developments in earth-based telescopes mean that they are catching up on Hubble, though Hubble still has some unique capabilities. But because the atmosphere absorbs IR, they are blind in that range. And there is just as much bandwidth and just as much interesting information out there in the IR. Probably more, because you can see things that are cooler and hence at an earlier stage in their development. Most of the pretty pictures which we see and enjoy so much are in false colour anyway.
Fine - provided you are willing to be limited to a vocabulary of 100-200 words, according to TFA.
Speech recognition for limited vocabularies has worked quite well for a long time, but quud quality real speech recognition is still over the horizon.I think this will go the same way.
No. The shockwave from the supernova produced localized density increases in a nearby or surrounding gas cloud. These density increases pushed the local gravitational field over the level at which the gas begins to accrete into what will eventually become a star. Such shock waves are the main cause of starts being formed, and the reason why there are "star nurseries" - volumes of space in which a large number of new stars are being born.
Not this (right) side of the atlantic, it doesn't. "Begging the question" is, and always has been, a precise term for a particular logical error. You are attempting to foorce a change in the language to match your own particular taste; I and others will fight back with what is not only historicalusage but our everyday usage. (Likewise with your usage of moot. Inflammable has never meant non-flammable; it means capable of being inflamed, rendered into flames. The in- prefix has two sources, one meaning not from Greek and one meaning into or towards from Latin - this is the latter).
Google should take them literally. Since Google makes money by selling ads on searches, Google sshould interpret their request as one to remove all their material from all their sites. This would opt them out ofg the worlds top search engine - at their own request. I wonder whether they would like becoming invisible to half the users on the Net.
I Engrish, but I rotten typist - see Sig. We need a syntax and semantic chacker for/. subhmissions. (I compiler would be too much to ask - you don't expect them to make sense, jsut to be valid.
If you could get the message to them and be believed in time, probably. The problem is that they have had a lifetime of being told thet Kim Jong Il and Kim Il Sung are the source of all things good, and thet everybody outside NK is a lying hyaena out to destroy the workers paradise. It wouldn't take many weeks to overturn this. The trouble is that Seoul is within easy artillery shot of the NK aremy, and it wouldn't take many hours for NK to wreck large areas of it and kill tens if not hundreds of thousands of S Koreans. Even malnourished soldiers (and the ones near the borders get the best food) could fire off a hell of a lot of shells before they were overrun.
The other problem is that if you do get through to them, every singlr North Korean is going to want out of there fast, and you will have 60 million refugees flooding into S Korea and China, or anywhere they can get a boat to. A problem that would make the Vietnamese Boat People look like a trickle. Both SK and China are terrified of this. China could probably topple the NK govenment within weeks if it wanted to - but it is desperately afraid of tha anarch that would follow. The same Economist article said that it was rumpured that the Chines army had stiied whether it was possible to take over NK "blizkrieg" style, so as to be in charge before the country collapsed into chaos, and had come to the conclusion that it was impossible.
According to this weeks Economist, even the army sometimes starves. Families fight to help therir sons avoid concsription into army units notorious for malnutrition.
Add another noisemaker at a slightly higher frequency, inaudible even to teenagers. They beat to create an unpleasant tone audible to all. When the original is turned off, unpleasant noise disappears. Proprietor of original thinks it doesn't work and disables it.
The first ever video tapes, the Ampex 2-inc machines, you could see the frames if you put a supension if iron filings on the tape. The video coule be edited with a sharp (non-magnetic) knofe by cutting between frames in the blanking and splicing the two pieces together, just like film.
It's just the total complexity of the system. Most successful systems are simple enought hat, at some level, one person (such as Woz) can understand the whole system; and the purts on which that system are well understood and well characterised. In the se of the Shuttle, there too many parts, and too many of the parts are designed for that system alone, for anyone to understand the whole thing.
Re:How about some more *durable* flash drives?
on
16GB Flash USB Dongle
·
· Score: 1
No - it depends upon how often you write to it. If you just fill it up incrementally with data, it will be very reliable. But if you put a diffeerent selection of movies on every weekend, you will eventually exceed the write cycles. And if you run an OS on ut, you will quickly exceed the write cycles.
YOu say your proposal would "put working-class people out of work". Round here it is difficult to get qualified bus drivers. One of the things which makes public transport round here unreliable, and less used, is that if a driver goes off sick, they have no replacement so that the bus doesn't run. So people dont't trust busses, and use cars instead. And truae automated drivers for urban streeds are a decade away at least (you might get automated drivers for multi-lane freeways sooner).
A possibly better scheme I saw proposed was based on car sharing using GPS enabled mobile phones. When you set out you set up your destination (as you can do now in your Sat nav). Likewise, someone wanting to make a journey without a car enters their destination into their mobile. A computer pairs the two and directs any driver with (say) less than 0.5 mile divert at each end to the wating passenger. The computer can also take into account preferences (e.g some women would want lifts with women only), record who travelled with who ijh case of complaint, and charge the passneger for the journey, forwarding it minus commission to the driver. Which gives you a generalised any-to-any taxi service with no extra mileage. Thos who have cars get costs reduced, those who haven't get taxi-quality transport.
When we didn't have cars, we could only take jobs, go to entertainments, visit friends which were either within walking/cycling distance, or which were in the public transport routes that ran near our house. If you wanted to take a new job somewhere else, you had to move house. But that was easier than now, becausee wives either didn't have jobs or had menial, non-career, jobs that they could drop and pick up again. And most houses were rental, and most people had fewer belongings to move.
Cars have given us freedom to choose jobs within, roughly, a 90 minute road commute - which can be a very large area. This is good for the worker - many more jobs to choose from, so you can optimise your choices. And good for companies too, for the same reason - they can pick the best workers for their needs rather than having to put up with the ones who live locally. And as the world has gained more and more different skill sets, that has become more important. When 90% of workers were semi- or un-skilled, they were more or less interchangeable: as long as there were 100 free workers in the area, ypou could find 50 thyat you need. But if you need one of the only 10 skilled flange-wobblers in your mega-city, they may have to travel a long way to your facility. Or move house - except that their spouse has a job where they live now, which brings in 50% of the household income.
I have experience of this as a governor of a specialist school, when we need to recruit new senior staff. Being a specialist school, there are not many about. Nearly all the applicants, and all the appointees, have had journeys of over 50 miles to the school, and non-moveable spouses. Without cars, we would have had to appoint inferior head teachers.
So we will not switch on a large scale to public transport for the trip to work unless we are willing to give up a freedom which most of us value highly - and one which has probably contributed to the economic growth of our countries. The correlation between wealth and number of cars runs a bit both ways: more wealth allows us to buy more cars, but more cars allow us to fine-tune our economy. When you are talking about annual growth rates of 2-3%, an extra 1% growth because you can place people better shows up. And remember this growth is compound.
So what ewe need is a "screen saver" - a program which runs in the processor idle state - which checks out as much of the processor as it is reasonably able to do. If that runs for a minute or two every hour, you're going to have a pretty high confidence that the CPU is working properly.
Correct. The size of the cells used by telephones varies enormously, and hence the power to cover the cell properly also varies. In crowded areas with heavy cellphone use, such as city financial centres, the cells may be only 100 yards across. The power is turned doen so as to avoid invading nearby cells. On the other hand, in isolated regions, they want to make a few masts cover as much area as possible, so they turn up the power so the cells may be tens of miles across. But whatever power you are using, you don't want to be heard across an adjacent cell - the ideal is a small overlap bbetween adjacent cells but no crosstalk to cells beyond. So both masts and phones continuously adjust their power to be "just right". The rain signal discussed in this article is basically the level of this adjustment.
You can bet that when a phon is advertised as having "up to 240 minutes talk time", that means you get that talk time when standing very close to the mast and therefore using minimum power. In real use, you will be further away, need more power, and get less talk time
Presumabley, because (as was said) they wer consideting dumping the UP kernel and shipping only the SMP kernel.
Obviously, SMP will haev some overhead - but what is that? 1% woulkd be so small that you could ignore it for nearly all purposes. 15% is too large to ignore. It seems to me a perfect reasonable piece of research to measure what that overhead actually is.
Not to belittle the energy savings, but how fast is it compared to a clocked CPU with a similar instruction set?
50-70MHz, so not very fast. This is not intended as a number-crunching engine, but as intelligence for vey low power devices.
That said, if this thing works properly in mass production, I would expect it to speed up very rapidly over the next few years. I bet they have used very conservative design rules in this first iteration, so as not to get the concept a bad reputation. Once they have the concept established, they will be able to move forward.
it has to be limited at least by the time it takes an electron to go from one end of the processor to another. That is about as fast as any processor can get.
No it doesn't. That is one of the advangages if a clockless design. Each stage presents the results to the next stage when they are ready. If the receiving stage is near the transmitting stage, then it gets on with calculating the results at once, rather than waiting, as it would have to do in a clocked design, until the signals would have reached a non-existent receiver on the other side of the chip.
The handshaking necessary to achieve the clockless design slows it down, but asynchronous designs have the potential to go very fast indeed. I would espect these designs to speed up as the engineers begin to work out how to shorten the physical paths between functional units. I also think that hyperthreading might bring much bigger returns on such processors, because threads could "wander" round the die without interfereing with each other.
There would not normally be any sector seeks in a consecutive burst. Any modern drive is capable of reading consecutive sectors (a) in one command and (b) in one pass of the oxide under the heads. 4096 data bytes is 32768 bits. With ECC etc, the total number of data bits which have to be transferred is probably about 38 k. If it were formatted as 8 512 byte sectors, they heads would also have to fly over 7 inter-block gaps, each probably a hundred or two bits long. Also, since there would seperate ECC blocks foir each 512 bytes instead of a single slightly larger one, there would be some wastage there. But the time taken would be about the time for the heads to fly over 40k raw bits instead of 38k raw bits - negligible. And the data storage wold go up comparably.
About 15 years ago, when the 3.5 in HDD held 500Mbyte, you could reformat your SCSI disk to get an optimim sector size. The disks I was using handled, I think, any even sector size from 128 bytes to 4096 bytes.
Because the disks were so low capacity, you wanted to use every byte, so I reformatted the disks to an optimum sector size for my application, which was about 1812 bytes IIRC. This achieved about 5% extra useful data on the system.
I think there was at one time a need for 2040 byte sectors for the IBM System/38, which had a 33rd bit on each word, which had to be saved to disk.
When the generation of disk changed to 1Mbyte, the controllers had an error one in every few tens of thousands of reads: it simply never completed the transaction. It never happened with 512 byte sectors, and the drive manufacturer only tested at 512 bytes, so we switched back to 512 and have been there ever since.
Slashdot Mirror
Almost certainly completely hidden by the driver interface. Firstly because it would be easy to do, secondly because I doubt that (e.g.) the SATA interface has commands to handle this new technology invented about eight years after it was designed, and thirdly because keeping it built-in vastly simplifies the power fail handling.
This is exactly how I thought that evolution was supposed to work. Environment changes due to new predator, species evolves to handle changed environment. Yet another routine confirmation of Darwin.
How are longer legs "behaviour"? Not that I would be surprised that behaviour has Darwinian consequences. The behaviour of reproductive mating has considerable consequences - species that stop reproducing lose the evolutionary challenge.
About as dumb an article as you get
So the Chinese and Taiwanese are dominating low-margin consumer goods, which you come across every day.
Now look where your high-margin, high tech gear is coming from. Who makes your aircraft, routers, heavy engineering equipment? The sort of thing you only see in a professional environment, and which has twenty times the markup of consumer goods. Overwhelmingly from the US, with Japan and the EU following behind.
China is doing unskilled and low skilled work. The "West" is still holding in there well on the lucrative tech stuff, and will continue to do so for some years yet. But don't rest on your laurels - if you stop running, others will overtake.
"Only" does infrared? Actually, there is more information to be gathered in the infrared than there is in the visible. Developments in earth-based telescopes mean that they are catching up on Hubble, though Hubble still has some unique capabilities. But because the atmosphere absorbs IR, they are blind in that range. And there is just as much bandwidth and just as much interesting information out there in the IR. Probably more, because you can see things that are cooler and hence at an earlier stage in their development. Most of the pretty pictures which we see and enjoy so much are in false colour anyway.
Fine - provided you are willing to be limited to a vocabulary of 100-200 words, according to TFA.
Speech recognition for limited vocabularies has worked quite well for a long time, but quud quality real speech recognition is still over the horizon.I think this will go the same way.
No. The shockwave from the supernova produced localized density increases in a nearby or surrounding gas cloud. These density increases pushed the local gravitational field over the level at which the gas begins to accrete into what will eventually become a star. Such shock waves are the main cause of starts being formed, and the reason why there are "star nurseries" - volumes of space in which a large number of new stars are being born.
Not this (right) side of the atlantic, it doesn't. "Begging the question" is, and always has been, a precise term for a particular logical error. You are attempting to foorce a change in the language to match your own particular taste; I and others will fight back with what is not only historicalusage but our everyday usage. (Likewise with your usage of moot. Inflammable has never meant non-flammable; it means capable of being inflamed, rendered into flames. The in- prefix has two sources, one meaning not from Greek and one meaning into or towards from Latin - this is the latter).
Google should take them literally. Since Google makes money by selling ads on searches, Google sshould interpret their request as one to remove all their material from all their sites. This would opt them out ofg the worlds top search engine - at their own request. I wonder whether they would like becoming invisible to half the users on the Net.
I Engrish, but I rotten typist - see Sig. We need a syntax and semantic chacker for /. subhmissions. (I compiler would be too much to ask - you don't expect them to make sense, jsut to be valid.
If you could get the message to them and be believed in time, probably. The problem is that they have had a lifetime of being told thet Kim Jong Il and Kim Il Sung are the source of all things good, and thet everybody outside NK is a lying hyaena out to destroy the workers paradise. It wouldn't take many weeks to overturn this. The trouble is that Seoul is within easy artillery shot of the NK aremy, and it wouldn't take many hours for NK to wreck large areas of it and kill tens if not hundreds of thousands of S Koreans. Even malnourished soldiers (and the ones near the borders get the best food) could fire off a hell of a lot of shells before they were overrun.
The other problem is that if you do get through to them, every singlr North Korean is going to want out of there fast, and you will have 60 million refugees flooding into S Korea and China, or anywhere they can get a boat to. A problem that would make the Vietnamese Boat People look like a trickle. Both SK and China are terrified of this. China could probably topple the NK govenment within weeks if it wanted to - but it is desperately afraid of tha anarch that would follow. The same Economist article said that it was rumpured that the Chines army had stiied whether it was possible to take over NK "blizkrieg" style, so as to be in charge before the country collapsed into chaos, and had come to the conclusion that it was impossible.
According to this weeks Economist, even the army sometimes starves. Families fight to help therir sons avoid concsription into army units notorious for malnutrition.
Add another noisemaker at a slightly higher frequency, inaudible even to teenagers. They beat to create an unpleasant tone audible to all. When the original is turned off, unpleasant noise disappears. Proprietor of original thinks it doesn't work and disables it.
The first ever video tapes, the Ampex 2-inc machines, you could see the frames if you put a supension if iron filings on the tape. The video coule be edited with a sharp (non-magnetic) knofe by cutting between frames in the blanking and splicing the two pieces together, just like film.
It's just the total complexity of the system. Most successful systems are simple enought hat, at some level, one person (such as Woz) can understand the whole system; and the purts on which that system are well understood and well characterised. In the se of the Shuttle, there too many parts, and too many of the parts are designed for that system alone, for anyone to understand the whole thing.
No - it depends upon how often you write to it. If you just fill it up incrementally with data, it will be very reliable. But if you put a diffeerent selection of movies on every weekend, you will eventually exceed the write cycles. And if you run an OS on ut, you will quickly exceed the write cycles.
YOu say your proposal would "put working-class people out of work". Round here it is difficult to get qualified bus drivers. One of the things which makes public transport round here unreliable, and less used, is that if a driver goes off sick, they have no replacement so that the bus doesn't run. So people dont't trust busses, and use cars instead. And truae automated drivers for urban streeds are a decade away at least (you might get automated drivers for multi-lane freeways sooner).
A possibly better scheme I saw proposed was based on car sharing using GPS enabled mobile phones. When you set out you set up your destination (as you can do now in your Sat nav). Likewise, someone wanting to make a journey without a car enters their destination into their mobile. A computer pairs the two and directs any driver with (say) less than 0.5 mile divert at each end to the wating passenger. The computer can also take into account preferences (e.g some women would want lifts with women only), record who travelled with who ijh case of complaint, and charge the passneger for the journey, forwarding it minus commission to the driver. Which gives you a generalised any-to-any taxi service with no extra mileage. Thos who have cars get costs reduced, those who haven't get taxi-quality transport.
When we didn't have cars, we could only take jobs, go to entertainments, visit friends which were either within walking/cycling distance, or which were in the public transport routes that ran near our house. If you wanted to take a new job somewhere else, you had to move house. But that was easier than now, becausee wives either didn't have jobs or had menial, non-career, jobs that they could drop and pick up again. And most houses were rental, and most people had fewer belongings to move.
Cars have given us freedom to choose jobs within, roughly, a 90 minute road commute - which can be a very large area. This is good for the worker - many more jobs to choose from, so you can optimise your choices. And good for companies too, for the same reason - they can pick the best workers for their needs rather than having to put up with the ones who live locally. And as the world has gained more and more different skill sets, that has become more important. When 90% of workers were semi- or un-skilled, they were more or less interchangeable: as long as there were 100 free workers in the area, ypou could find 50 thyat you need. But if you need one of the only 10 skilled flange-wobblers in your mega-city, they may have to travel a long way to your facility. Or move house - except that their spouse has a job where they live now, which brings in 50% of the household income.
I have experience of this as a governor of a specialist school, when we need to recruit new senior staff. Being a specialist school, there are not many about. Nearly all the applicants, and all the appointees, have had journeys of over 50 miles to the school, and non-moveable spouses. Without cars, we would have had to appoint inferior head teachers.
So we will not switch on a large scale to public transport for the trip to work unless we are willing to give up a freedom which most of us value highly - and one which has probably contributed to the economic growth of our countries. The correlation between wealth and number of cars runs a bit both ways: more wealth allows us to buy more cars, but more cars allow us to fine-tune our economy. When you are talking about annual growth rates of 2-3%, an extra 1% growth because you can place people better shows up. And remember this growth is compound.
So what ewe need is a "screen saver" - a program which runs in the processor idle state - which checks out as much of the processor as it is reasonably able to do. If that runs for a minute or two every hour, you're going to have a pretty high confidence that the CPU is working properly.
Correct. The size of the cells used by telephones varies enormously, and hence the power to cover the cell properly also varies. In crowded areas with heavy cellphone use, such as city financial centres, the cells may be only 100 yards across. The power is turned doen so as to avoid invading nearby cells. On the other hand, in isolated regions, they want to make a few masts cover as much area as possible, so they turn up the power so the cells may be tens of miles across. But whatever power you are using, you don't want to be heard across an adjacent cell - the ideal is a small overlap bbetween adjacent cells but no crosstalk to cells beyond. So both masts and phones continuously adjust their power to be "just right". The rain signal discussed in this article is basically the level of this adjustment.
You can bet that when a phon is advertised as having "up to 240 minutes talk time", that means you get that talk time when standing very close to the mast and therefore using minimum power. In real use, you will be further away, need more power, and get less talk time
Presumabley, because (as was said) they wer consideting dumping the UP kernel and shipping only the SMP kernel.
Obviously, SMP will haev some overhead - but what is that? 1% woulkd be so small that you could ignore it for nearly all purposes. 15% is too large to ignore. It seems to me a perfect reasonable piece of research to measure what that overhead actually is.
Not to belittle the energy savings, but how fast is it compared to a clocked CPU with a similar instruction set?
50-70MHz, so not very fast. This is not intended as a number-crunching engine, but as intelligence for vey low power devices.
That said, if this thing works properly in mass production, I would expect it to speed up very rapidly over the next few years. I bet they have used very conservative design rules in this first iteration, so as not to get the concept a bad reputation. Once they have the concept established, they will be able to move forward.
it has to be limited at least by the time it takes an electron to go from one end of the processor to another. That is about as fast as any processor can get.
No it doesn't. That is one of the advangages if a clockless design. Each stage presents the results to the next stage when they are ready. If the receiving stage is near the transmitting stage, then it gets on with calculating the results at once, rather than waiting, as it would have to do in a clocked design, until the signals would have reached a non-existent receiver on the other side of the chip.
The handshaking necessary to achieve the clockless design slows it down, but asynchronous designs have the potential to go very fast indeed. I would espect these designs to speed up as the engineers begin to work out how to shorten the physical paths between functional units. I also think that hyperthreading might bring much bigger returns on such processors, because threads could "wander" round the die without interfereing with each other.
There would not normally be any sector seeks in a consecutive burst. Any modern drive is capable of reading consecutive sectors (a) in one command and (b) in one pass of the oxide under the heads. 4096 data bytes is 32768 bits. With ECC etc, the total number of data bits which have to be transferred is probably about 38 k. If it were formatted as 8 512 byte sectors, they heads would also have to fly over 7 inter-block gaps, each probably a hundred or two bits long. Also, since there would seperate ECC blocks foir each 512 bytes instead of a single slightly larger one, there would be some wastage there. But the time taken would be about the time for the heads to fly over 40k raw bits instead of 38k raw bits - negligible. And the data storage wold go up comparably.
About 15 years ago, when the 3.5 in HDD held 500Mbyte, you could reformat your SCSI disk to get an optimim sector size. The disks I was using handled, I think, any even sector size from 128 bytes to 4096 bytes.
Because the disks were so low capacity, you wanted to use every byte, so I reformatted the disks to an optimum sector size for my application, which was about 1812 bytes IIRC. This achieved about 5% extra useful data on the system.
I think there was at one time a need for 2040 byte sectors for the IBM System/38, which had a 33rd bit on each word, which had to be saved to disk.
When the generation of disk changed to 1Mbyte, the controllers had an error one in every few tens of thousands of reads: it simply never completed the transaction. It never happened with 512 byte sectors, and the drive manufacturer only tested at 512 bytes, so we switched back to 512 and have been there ever since.
Mod parent up please, somebody. That is a very insightful article.