Believe it or not, but Google already does this extremely well. If you use several android devices and chrome on your computers you'll start to notice that the search results always become more and more personalised to your needs. Obviously this does require you to login (something many people forget to do).
If it's just a program that decides what large blocks you should put in a SoC then it's just plain stupid really. The first step to proper design is hiring engineers who aren't retarded monkeys that are able to make such design decisions easily.
Any mistake in a SoC is expensive, especially if you go directly from design to wafer without extensive testing. Most of the time it's 1 week of actual writing a description in VHDL or Verilog and then spending a few weeks/months verifying the design and removing any bug. After you're done verifying you get to enjoy the trouble of mapping it. And since automation sucks for certain things, especially when the analogue signals are supposed to remain as noise free as possible, you then get to spend a few days doing the layout and verifying if the layout is actually the same as the netlist.
While this sounds nice, it's not the first "C to Silicon" program out there (Cadence beat them to it). And it certainly won't be the last. The thing is, the reason to use VHDL and to some extent Verilog is to minimize the occurrence of errors. Even when you verify your design bugs can still slip through. But due to the overall design of the language this is far less likely in VHDL.
Wrong though; units don't scale like that. It's funny how people use an average for one thing and then assume it expands to an entire set without having to modify anything.
Look at the atmospheric concentration of methane particles vs. carbon dioxide particles. Next look at how much percentage of the greenhouse effect it accounts for.
Actually methane levels went up 50 ppb in the past decade... Might not sound much until you realize that's an increase that's coming close to 3% increase in a single decade. Units are very important;)
At the current increasing rate of methane output it's actually a larger danger than carbon dioxide. Additionally in all your carbon dioxide hyping you forgot that an increase in temperatures (due to methane among other things) actually causes for an increase in water vapor in the atmosphere...
It does so very much. Learn to do basic math. One particle of methane has 200 times more effect than one particle of carbon dioxide. Additionally one particle of methane added to the total amount of methane is far more significant than one particle of carbon dioxide being added.
Actually, a gas with a short half life that's a lot more effective but not being looked after due to some environmentalist hype is far more dangerous than a gas with a long half life with minor effects. As it will cause an increase in water vapor concentration (due to increasing temperature), resulting in an even stronger greenhouse effect. Cause yes, water vapor is in fact a greenhouse gas.
Look into what greenhouse gas is the worst per unit of concentration. You'll find that carbon dioxide isn't even worth worrying about. Methane is our real problem, and nobody is trying to lower its emissions at all while its effect per unit of concentration far exceeds that of carbon dioxide.
You might want to take a look at methane. You might want to compare the percentage in the total effect versus the concentration. It contributes about 10% of the greenhouse effect while it has a concentration that's roughly 200 times lower than carbon dioxide. And carbon dioxide only accounts for roughly 30% of the greenhouse effect. So even a minor increase in the concentration of methane has a far larger effect than that of carbon dioxide. And there is almost no effort to stop the emissions of methane compared to those of carbon dioxide.
They are correct though. There are way worse greenhouse gasses that don't even get filtered most of the time. Cause actually carbon dioxide isn't all that strong of a greenhouse gas.
That's where you're wrong, the cost per area is actually increasing significantly. It is indeed true that the cost per transistor is decreasing (at the moment at least). But since for example Intel and AMD want performance, so they're willing to trade in significant portions of area for an increase in speed. So the cost of your average desktop processor should actually keep increasing; something that luckily hasn't been happening considerably compared to the increasing costs of the new lithography machines and masks. Energy consumption is only a secondary concern when performance is the main factor in most cases as long as it's possible to evacuate the thermal energy and it doesn't have to work off a battery.
The number you're looking for is somewhere between 30 and 40% actually for most processors. But an additional factor is that memory is a regular structure, this means you can open up the lithography trick book. This allows for a significant decrease in size compared to regular logic. But this not being very relevant to the original subject I'll shut up about it. The actual issue is that you often have to wait for data, this is the major bottleneck of any processor (not to mention it can stall the pipeline). This is why DMA is so interesting, the waiting times are kept to a minimum. On the other hand if you work on a packet based network you're in for quite some problems. You need a lot of additional logic to manage this new bus. Logic that is very expensive in terms of silicon area. And it's questionable EUV lithography will be ready in time to save our asses on this one, ASML says it'll still take a while and I think we can take their word on it. Another concern is the length of the logic chain causing even longer propagation delays.
To keep it short, a more sophisticated network is indeed necessary. But just saying packet based will fix it is plain wrong. What I do see happening is groups of 4 or 16 cores being interconnected using a more sophisticated network, but even then I question if it'll be packet based due to timing issues.
That's just plain inefficient use of silicon area. They wish to waste some of that limited space on additional logic that isn't strictly necessary. And it will cause a significant bottleneck to be created. Did they forget about DMA controllers or something? You already need a DMA controller no mater what and it's perfectly capable of accessing the necessary memories as it is. Adding some extra capabilities to the DMA controller would be far more efficient in logic area size and most likely lead to a better performance compared to this bad idea.
If necessary you can use a FPGA for decoding the video and audio stream. Even a cheap, small FPGA (maybe with a DSP processor as co-processor to assist it) will easily be able to handle a H.264 HD stream in real time at fairly low clock speeds. The issue is the logic size for multi format support. This could be solved by having the ARM processor load the necessary FPGA configuration on the fly. And having the ARM only generate the GUI. The moment playback starts control over the display and video source is handed over to the FPGA. This would be somewhat complicated but it'd certainly be an interesting design exercise.
The beagle board has been around a lot longer, has more features and is very well documented. Additionally the ARM processor on the beagle board is more powerful and actually intended for multimedia and signal processing applications. Do some research into what was first before you wish to make comments like this.
A good ARM board with proper multimedia functionality should be sufficient (I think the Beagle board might be sufficient). Though obviously the larger FPGAs would excel at this. But it'd take quite a lot of time to rewrite everything in VHDL or Verilog. And even then, you'd need one of the larger more expensive FPGA's with enough slices. In the end it'd be easier to grab an old computer and make your own IR sensor and use one of those universal remotes with it.
The first use will be warfare as is always the case sadly. You'll probably first see rail- and coil-guns show up. Next you'll find its uses in radars and specifically in trying to make them useless. Then it will proceed into gimmicks for rich people. After that it'll go to civil scientists (space exploration, particle accelerators,...) and maybe a few years later into people's houses. Somewhere in between all of that somebody might find a use for it in medicine (other than improving your standard NMRI).
Here we go, again another round of 1980 hysteria in the "government is going to track our every move" category.
First of all, if they wanted to track you; then they'd sure as hell do it no matter what. They don't need high tech gadgets to do that either.
But on to RFID. Our student ID cards at college/university here work by RFID. We use them to open doors, as copy cards, to pay for food in the cafeteria... . So you end up with a bunch of (electronic/electrical) engineering students with RFID cards that you actually get to store money on indirectly (its kept on the server how much money is on the card). So there are enough reasons to duplicate somebody else's card. Contrary to what you might be thinking, not a single person has even managed to do that successfully. The encoding is fairly tricky and just replicating it is extremely hard.
What we did try was to track each other using the RFID tags in these cards (as that's considerably easier than trying to reverse engineer them). The thing is, a tracking range of a whole 15cm isn't all that useful last time I checked. So claiming you can track people through these tags is foolish. Unless if they're passing through lets say a door frame.
So I seriously doubt that any high school kid will be able to figure it out that easily, all this does is save the school some time in the morning by being able to skip the daily roll call.
Slashdot Mirror
Believe it or not, but Google already does this extremely well. If you use several android devices and chrome on your computers you'll start to notice that the search results always become more and more personalised to your needs. Obviously this does require you to login (something many people forget to do).
Verilog is still quite different from C in many aspects at first glance. I'm not really familiar with Verilog to be honest, so might be mistaken.
If it's just a program that decides what large blocks you should put in a SoC then it's just plain stupid really. The first step to proper design is hiring engineers who aren't retarded monkeys that are able to make such design decisions easily.
Any mistake in a SoC is expensive, especially if you go directly from design to wafer without extensive testing. Most of the time it's 1 week of actual writing a description in VHDL or Verilog and then spending a few weeks/months verifying the design and removing any bug. After you're done verifying you get to enjoy the trouble of mapping it. And since automation sucks for certain things, especially when the analogue signals are supposed to remain as noise free as possible, you then get to spend a few days doing the layout and verifying if the layout is actually the same as the netlist.
While this sounds nice, it's not the first "C to Silicon" program out there (Cadence beat them to it). And it certainly won't be the last. The thing is, the reason to use VHDL and to some extent Verilog is to minimize the occurrence of errors. Even when you verify your design bugs can still slip through. But due to the overall design of the language this is far less likely in VHDL.
Wrong though; units don't scale like that. It's funny how people use an average for one thing and then assume it expands to an entire set without having to modify anything.
Look at the atmospheric concentration of methane particles vs. carbon dioxide particles. Next look at how much percentage of the greenhouse effect it accounts for.
Actually there were several periods in past centuries where it was warmer on average than it is now...
Actually methane levels went up 50 ppb in the past decade... Might not sound much until you realize that's an increase that's coming close to 3% increase in a single decade. Units are very important ;)
No it doesn't get obliterated. 1/200th of the quantity of CO2 produces one third of its effect. I call that pretty significant.
At the current increasing rate of methane output it's actually a larger danger than carbon dioxide. Additionally in all your carbon dioxide hyping you forgot that an increase in temperatures (due to methane among other things) actually causes for an increase in water vapor in the atmosphere...
Yes, and methane causes an increase in that temperature. Learn to read...
It does so very much. Learn to do basic math. One particle of methane has 200 times more effect than one particle of carbon dioxide. Additionally one particle of methane added to the total amount of methane is far more significant than one particle of carbon dioxide being added.
Actually, a gas with a short half life that's a lot more effective but not being looked after due to some environmentalist hype is far more dangerous than a gas with a long half life with minor effects. As it will cause an increase in water vapor concentration (due to increasing temperature), resulting in an even stronger greenhouse effect. Cause yes, water vapor is in fact a greenhouse gas.
Would you please stop reading things that aren't there? Where did I claim that fossil fuels have nothing to do with methane emissions?
Look into what greenhouse gas is the worst per unit of concentration. You'll find that carbon dioxide isn't even worth worrying about. Methane is our real problem, and nobody is trying to lower its emissions at all while its effect per unit of concentration far exceeds that of carbon dioxide.
You might want to take a look at methane. You might want to compare the percentage in the total effect versus the concentration. It contributes about 10% of the greenhouse effect while it has a concentration that's roughly 200 times lower than carbon dioxide. And carbon dioxide only accounts for roughly 30% of the greenhouse effect. So even a minor increase in the concentration of methane has a far larger effect than that of carbon dioxide. And there is almost no effort to stop the emissions of methane compared to those of carbon dioxide.
They are correct though. There are way worse greenhouse gasses that don't even get filtered most of the time. Cause actually carbon dioxide isn't all that strong of a greenhouse gas.
That's where you're wrong, the cost per area is actually increasing significantly. It is indeed true that the cost per transistor is decreasing (at the moment at least). But since for example Intel and AMD want performance, so they're willing to trade in significant portions of area for an increase in speed. So the cost of your average desktop processor should actually keep increasing; something that luckily hasn't been happening considerably compared to the increasing costs of the new lithography machines and masks. Energy consumption is only a secondary concern when performance is the main factor in most cases as long as it's possible to evacuate the thermal energy and it doesn't have to work off a battery.
The number you're looking for is somewhere between 30 and 40% actually for most processors. But an additional factor is that memory is a regular structure, this means you can open up the lithography trick book. This allows for a significant decrease in size compared to regular logic. But this not being very relevant to the original subject I'll shut up about it. The actual issue is that you often have to wait for data, this is the major bottleneck of any processor (not to mention it can stall the pipeline). This is why DMA is so interesting, the waiting times are kept to a minimum. On the other hand if you work on a packet based network you're in for quite some problems. You need a lot of additional logic to manage this new bus. Logic that is very expensive in terms of silicon area. And it's questionable EUV lithography will be ready in time to save our asses on this one, ASML says it'll still take a while and I think we can take their word on it. Another concern is the length of the logic chain causing even longer propagation delays.
To keep it short, a more sophisticated network is indeed necessary. But just saying packet based will fix it is plain wrong. What I do see happening is groups of 4 or 16 cores being interconnected using a more sophisticated network, but even then I question if it'll be packet based due to timing issues.
That's just plain inefficient use of silicon area. They wish to waste some of that limited space on additional logic that isn't strictly necessary. And it will cause a significant bottleneck to be created. Did they forget about DMA controllers or something? You already need a DMA controller no mater what and it's perfectly capable of accessing the necessary memories as it is. Adding some extra capabilities to the DMA controller would be far more efficient in logic area size and most likely lead to a better performance compared to this bad idea.
If necessary you can use a FPGA for decoding the video and audio stream. Even a cheap, small FPGA (maybe with a DSP processor as co-processor to assist it) will easily be able to handle a H.264 HD stream in real time at fairly low clock speeds. The issue is the logic size for multi format support. This could be solved by having the ARM processor load the necessary FPGA configuration on the fly. And having the ARM only generate the GUI. The moment playback starts control over the display and video source is handed over to the FPGA. This would be somewhat complicated but it'd certainly be an interesting design exercise.
The beagle board has been around a lot longer, has more features and is very well documented. Additionally the ARM processor on the beagle board is more powerful and actually intended for multimedia and signal processing applications. Do some research into what was first before you wish to make comments like this.
A good ARM board with proper multimedia functionality should be sufficient (I think the Beagle board might be sufficient). Though obviously the larger FPGAs would excel at this. But it'd take quite a lot of time to rewrite everything in VHDL or Verilog. And even then, you'd need one of the larger more expensive FPGA's with enough slices. In the end it'd be easier to grab an old computer and make your own IR sensor and use one of those universal remotes with it.
If they wanted to track everyone they'd still just do it. But they don't. Believe it or not, but you are too boring to be worth tracking.
The first use will be warfare as is always the case sadly. You'll probably first see rail- and coil-guns show up. Next you'll find its uses in radars and specifically in trying to make them useless. Then it will proceed into gimmicks for rich people. After that it'll go to civil scientists (space exploration, particle accelerators, ...) and maybe a few years later into people's houses. Somewhere in between all of that somebody might find a use for it in medicine (other than improving your standard NMRI).
Here we go, again another round of 1980 hysteria in the "government is going to track our every move" category.
First of all, if they wanted to track you; then they'd sure as hell do it no matter what. They don't need high tech gadgets to do that either.
But on to RFID. Our student ID cards at college/university here work by RFID. We use them to open doors, as copy cards, to pay for food in the cafeteria... . So you end up with a bunch of (electronic/electrical) engineering students with RFID cards that you actually get to store money on indirectly (its kept on the server how much money is on the card). So there are enough reasons to duplicate somebody else's card. Contrary to what you might be thinking, not a single person has even managed to do that successfully. The encoding is fairly tricky and just replicating it is extremely hard.
What we did try was to track each other using the RFID tags in these cards (as that's considerably easier than trying to reverse engineer them). The thing is, a tracking range of a whole 15cm isn't all that useful last time I checked. So claiming you can track people through these tags is foolish. Unless if they're passing through lets say a door frame.
So I seriously doubt that any high school kid will be able to figure it out that easily, all this does is save the school some time in the morning by being able to skip the daily roll call.