Intel is just trying to solve the power distribution issue, not eliminating the main down-conversion stage, which will *always* be external.
Maybe the first stage can still be integrated into the package if not in the die. Only putting the transistors into the package will allow them to dissipate in a far better way than is most existing motherboard designs. Some high end server motherboards and some high end GPU cards uses state of the art switching regulators with high frequency, small magnetic and small ceramic capacitors. There small size can possibly make realistic an integration into the package.
Now just stack the SDRAM as planned and enjoy a package with almost only I/O signals...
Or they are finding those Chinese made MB motherboards have such poor regulators, that they have to do some final regulation on the die to keep things stable?
Mod parent up.
This is a very interesting argument, especially now that chip require multiple voltages with dynamic setting and precise tolerances.
In my experience the capacitors and transistors are the two parts that can die early, mostly because of thermal stress, due to slow chemical change of some of there properties: the resistance tend to increase and the capacity tend to decrease. I really don't know anything about the chemical stability of the components integrated into the die, but I can see some interesting points: 1) The capacitors are composed of the metal layers of the die. There is no electrolyte involved, so there must be very reliable, like ceramic capacitors. 2) The transistors will probably be cooled more efficiently than in most actual motherboard.
So there is possibility that this new design will increase the reliability. This argument is still speculation as there likely rely on a first stage standard switching power supply to lower the voltage at an acceptable level for the die...
Bollocks! Since the internal VR uses the same process as the CPU itself, it can't sustain high input voltages, therefore a one-stage 12V to 0.9V conversion is just a pipe dream.
The longer pdf presentation actually shows the motherboard-level 12V to 2.2V VR, which would be still rated for the full power (85W plus margin). OTOH, it's quite impressive that the 22nm process has support for 2.2V CMOS.
True, but it is hard as the radio in a phone tends not to be open software.
Yes it's hard, but not impossible.
A USRP would be much better but then you need amplification and power. You are inside a metal tube and you need to get inside an antenna on the outside which is designed to go off when it receives a burst from a 50KW radar.
ACARS transmission are handled by the VHF Ground Station (VGS), not by the radar ! VGS seem to be in the 50W range only, according to this specification: http://www.selexelsag.com/internet/localization/IPC/media/docs/OTED100Radio-MGS100.pdf . So, a close sub-watt transmitter will certainly get received by the aircraft. I can for example add fake messages that will not cam from the VGS.
The transponder squirts data back using something like 20w or so. It would be hard to overwhelm that from inside the plane.
The downlink is likely not required to be hacked to abuse a vulnerability of the aircraft itself.
Inside the plane, RF goes by coax. Data goes by different means, usually twisted pair. Either way, the data wiring goes from the front-end in the cockpit to the avionics bay which is located underneath the cockpit (so no long cable runs).
Any non properly shielded cable can be a path for some RF signal, the original signal inside the cable do not matter. I have see device that are so sensible that there are able to receive transmission from the power supply wire ! Even an unrelated signal wire can act as a path for a RF signal.
Please remember that planes ship with standard flight control systems only. Cockpits and avionics are selected by airlines based on different options. It would be quite hard to try out every variant.
Large fleet likely use a few standard variant.
However flight test has a big increase in general RF "crud" in the fuselage as you have multiple high performance logging and telemeter systems with cabling all over the place.
Nope ? How can you be so certain ? I have passed long days in electromagnetic testing room, and I can say that you will be surprised by what can happens with complex and highly programmable electronics !
Your "demonstration" prove that a software modification can open up the frequency range. This is not a surprise as most RF subsystem uses DSP and only a minimal of analog components. Your example is just about a bug; think of what can happens if the DSP are fully reprogrammed... Yes, the signal can be weak, but it could be enough to deal with very sensible antenna of some receiver on the aircraft. Placing in a cleaver way some metal part can greatly increase the gain. The size of smartphone antenna will almost not see the fuselage if you put it close to a window. In addition, a lot of wires in a aircraft are not shielded to reduce weight, making path up to more sensible material.
If you have read the publication subject of this article, you will see that aircraft manufacturers have actually not worried at all about vulnerability. There is simply no protection at all in many protocols. Finally the current testing practice of device against electromagnetic field susceptibility is only done using a high energy sinus weave slowly changing his frequency. RF subsystem can communicate with fare fare lower energy than the tested limits if modulated in a proper way. Don't mix security and vulnerability !
Sorry, but to have a android device that can transmit and receive ACARS is close to impossible.
I would not bet on that !
The lasts superphones embeds so much high speed subsystems (2,3,4G/WiFia,b,n,g/BT/FM/AM/NFC/RFID/PAL/NTSC/HDMI/USB2,3/Audio/ and certainly a few more at each generation) that there are probably capable of processing some signals at virtually any frequencies if some high skilled hackers are motivated to do it.
Analog filters never cut abruptly; DSP can be reprogrammed to abuse the surrounding components. Any interfaces can leaks some creative signals. Take a look at this for example: http://bellard.org/dvbt/
Aside of the strict legal blablabla, the future will be limited by the increase of the cost of controlling and hunting the copy. Because the privacy will always be a sensitive subject, there is no way for a simple system to control any possible copy. The situation will be more and more complex, and so will be the tools that try to search something. And complex tools cots a lot. Who will pay for it ? Yes, the customer. At one point at one point of cost, others business models will be more profitable for the artists than obliterate a lot of money in making potential customers angry.
And WP was doomed before Nokia chose him, just read the market share progression. All the Microsoft great promise of a brighter future with WP7, WP7.5, WP8 tuned to be just big lies: the market share simply continued to decrease.
The drama is that Nokia was in the way of fixing there problems, allowing transition from Symbian to Meego with Qt, when the new CEO destroyed almost anything. At the time of this disruptive action, there was a lot of speculation about how a quick ans total switch to WP could possibly success. Now, two years later, WP7, WP7.5, WP8, and almost WP7.8 later, and 13 Lumia phones later, the ridiculous 2.5% market share very clearly show how big was the mistake. Not counting the manufacturers that have recently reduced if not throw away there WP product range.
This is in fact the real questions: * How the Nokia board slowly changed to the point it elected a CEO that bring Nokia to his fall. * How can the board allow disruptive and destructive action from the CEO without limitation or even a reaction ?
I wonder if Debian will not have the advantage to split the concept of release in different layers, like for example base, console-apps, desktop, graph-apps, etc... Upgrading the desktop layer should not affect too much the base and the graph-apps.
And the be competitive, Debian should probably propose more advanced revision of a package that the stick stable on to gave to the user more choice. I suspect that many enjoy a high stability but would like to install the last release of a small number of packages that bring important features to them. Yes there is backport, but it's a big mess to setup. The choice of the revision should be really simple from application like synaptic, apt-get, aptitude, etc...
Yes, long release cycle is not adapted to the today's fast moving projects. Stability is still a valuable goal, but there is not point in fixing too old revisions that users will not use and that upstream will nor car of anymore.
Aside of that, I found Debian still a very important project because it bring to the community a lot of very good and clean technologies. In this regard the multiarch will certainly be the beginning of a new area, especially now that a lot of different arch are coming to the mass in the form of almost anything but traditional PC.
The chance seem to be very small, but without that Wheezy will look like a old duck with his Gnome 3.4 when it will be released.
I hope that the Debian team will be cleaver enough to understand the advantage of providing a good classic desktop experience for people that will upgrade from Squeeze to Wheezy (I have tried Gnome 3.x and Unity and found them unproductive).
The idea they are a patent troll is completely and 100% absurd.
If you listen to a large amount of comments about Nokia this last two years, you will probably notice a dramatic increase of opinions that there is a lot of absurdities in the actual Nokia. The most absurd one is still that Nokia blindly ignore this reality...
Nokia is not a patent troll by any reasonable definition.
This was certainly true for a long part of the Nokia history. But the actual Nokia is something that have lost an extremely large amount of connections with the Nokia "mobile phone world leader" of the past. We are now forced to take notice that the actual Nokia is more and more close to the definition of patent troll. The latest new just confirm this trend.
And if you take in account the risk perception by the population ? I think that you underestimate how much emotional can be the decision to live near a nuclear reactor with your children. And you cannot avoid emotional decision in a human community. The basis of this emotion is not only based on the number of dead, but also to the consequence on the survivors. In the two major civil nuclear accident, you have to add that the authority have massively lied about the scale of the problem. How can you expect that now people will believe pro nuclear saying that it's now under control ? There said that since the first civil nuclear plant and have been proved false at too many occasion.
Aircraft and cars for example kill a number of people each year, but even if the former is massively safer than the second, the risk perception is not necessary lower for the aircraft.
Sure that the amount of radioactive material released by Tchernobyl and Fukushima is far below the 4.2 million m3, but there will last far far longer and have been disseminate without bound in the environment. The consequence is that it's difficult to predict where there will be concentrated, mostly by organic processing, but not only. There is report that some new buildings contain irradiated material for example.
There are actually more like hazardous stacks of radioactive wreaks that need constant attention. A truly decommissioned plant is safe and can be converted into an other uses.
Yes, there isn't any safe type of power generation, but the high concentration of long half live highly radioactive isotopes make the nuclear generation in a category of his own regarding long term risk.
Slashdot Mirror
Mod parent up.
This will also make the whole system more reliable to cheap off-chip regulators.
Intel is just trying to solve the power distribution issue, not eliminating the main down-conversion stage, which will *always* be external.
Maybe the first stage can still be integrated into the package if not in the die. Only putting the transistors into the package will allow them to dissipate in a far better way than is most existing motherboard designs. Some high end server motherboards and some high end GPU cards uses state of the art switching regulators with high frequency, small magnetic and small ceramic capacitors. There small size can possibly make realistic an integration into the package.
Now just stack the SDRAM as planned and enjoy a package with almost only I/O signals...
An on-die power controller still needs to have external capacitors, especially at the power levels we're talking about.
From the presentation PDF, the capacitors are builds from the the metal layers of the die.
Or they are finding those Chinese made MB motherboards have such poor regulators, that they have to do some final regulation on the die to keep things stable?
Mod parent up.
This is a very interesting argument, especially now that chip require multiple voltages with dynamic setting and precise tolerances.
Voltage regulators fail often.
And you want to replace them with what exactly ?
In my experience the capacitors and transistors are the two parts that can die early, mostly because of thermal stress, due to slow chemical change of some of there properties: the resistance tend to increase and the capacity tend to decrease. I really don't know anything about the chemical stability of the components integrated into the die, but I can see some interesting points: 1) The capacitors are composed of the metal layers of the die. There is no electrolyte involved, so there must be very reliable, like ceramic capacitors. 2) The transistors will probably be cooled more efficiently than in most actual motherboard.
So there is possibility that this new design will increase the reliability. This argument is still speculation as there likely rely on a first stage standard switching power supply to lower the voltage at an acceptable level for the die...
Bollocks! Since the internal VR uses the same process as the CPU itself, it can't sustain high input voltages, therefore a one-stage 12V to 0.9V conversion is just a pipe dream.
The longer pdf presentation actually shows the motherboard-level 12V to 2.2V VR, which would be still rated for the full power (85W plus margin). OTOH, it's quite impressive that the 22nm process has support for 2.2V CMOS.
The actual FIVR don't use the same process as the CPU. It's a separete die using a 90nm process. Read the page 7 of http://www.psma.com/sites/default/files/uploads/tech-forums-nanotechnology/resources/400a-fully-integrated-silicon-voltage-regulator.pdf for the details.
Keep in mind that the Intel solution use a 30MHz to 140MHz switching frequency, compared to the 0.5MHz of the LTM4620.
True, but it is hard as the radio in a phone tends not to be open software.
Yes it's hard, but not impossible.
A USRP would be much better but then you need amplification and power. You are inside a metal tube and you need to get inside an antenna on the outside which is designed to go off when it receives a burst from a 50KW radar.
ACARS transmission are handled by the VHF Ground Station (VGS), not by the radar ! VGS seem to be in the 50W range only, according to this specification:
http://www.selexelsag.com/internet/localization/IPC/media/docs/OTED100Radio-MGS100.pdf . So, a close sub-watt transmitter will certainly get received by the aircraft. I can for example add fake messages that will not cam from the VGS.
The transponder squirts data back using something like 20w or so. It would be hard to overwhelm that from inside the plane.
The downlink is likely not required to be hacked to abuse a vulnerability of the aircraft itself.
Inside the plane, RF goes by coax. Data goes by different means, usually twisted pair. Either way, the data wiring goes from the front-end in the cockpit to the avionics bay which is located underneath the cockpit (so no long cable runs).
Any non properly shielded cable can be a path for some RF signal, the original signal inside the cable do not matter. I have see device that are so sensible that there are able to receive transmission from the power supply wire ! Even an unrelated signal wire can act as a path for a RF signal.
Please remember that planes ship with standard flight control systems only. Cockpits and avionics are selected by airlines based on different options. It would be quite hard to try out every variant.
Large fleet likely use a few standard variant.
However flight test has a big increase in general RF "crud" in the fuselage as you have multiple high performance logging and telemeter systems with cabling all over the place.
Agree
Nope ? How can you be so certain ? I have passed long days in electromagnetic testing room, and I can say that you will be surprised by what can happens with complex and highly programmable electronics !
Your "demonstration" prove that a software modification can open up the frequency range. This is not a surprise as most RF subsystem uses DSP and only a minimal of analog components. Your example is just about a bug; think of what can happens if the DSP are fully reprogrammed... Yes, the signal can be weak, but it could be enough to deal with very sensible antenna of some receiver on the aircraft. Placing in a cleaver way some metal part can greatly increase the gain. The size of smartphone antenna will almost not see the fuselage if you put it close to a window. In addition, a lot of wires in a aircraft are not shielded to reduce weight, making path up to more sensible material.
If you have read the publication subject of this article, you will see that aircraft manufacturers have actually not worried at all about vulnerability. There is simply no protection at all in many protocols. Finally the current testing practice of device against electromagnetic field susceptibility is only done using a high energy sinus weave slowly changing his frequency. RF subsystem can communicate with fare fare lower energy than the tested limits if modulated in a proper way. Don't mix security and vulnerability !
Sorry, but to have a android device that can transmit and receive ACARS is close to impossible.
I would not bet on that !
The lasts superphones embeds so much high speed subsystems (2,3,4G/WiFia,b,n,g/BT/FM/AM/NFC/RFID/PAL/NTSC/HDMI/USB2,3/Audio/ and certainly a few more at each generation) that there are probably capable of processing some signals at virtually any frequencies if some high skilled hackers are motivated to do it.
Analog filters never cut abruptly; DSP can be reprogrammed to abuse the surrounding components. Any interfaces can leaks some creative signals. Take a look at this for example: http://bellard.org/dvbt/
... the basic desktop configurations that GNOME2 does like multi-head-separate-screen.
Ca you please provide a link that describes this feature ?
Aside of the strict legal blablabla, the future will be limited by the increase of the cost of controlling and hunting the copy. Because the privacy will always be a sensitive subject, there is no way for a simple system to control any possible copy. The situation will be more and more complex, and so will be the tools that try to search something. And complex tools cots a lot. Who will pay for it ? Yes, the customer. At one point at one point of cost, others business models will be more profitable for the artists than obliterate a lot of money in making potential customers angry.
He was doomed before the board chose him.
And WP was doomed before Nokia chose him, just read the market share progression. All the Microsoft great promise of a brighter future with WP7, WP7.5, WP8 tuned to be just big lies: the market share simply continued to decrease.
The drama is that Nokia was in the way of fixing there problems, allowing transition from Symbian to Meego with Qt, when the new CEO destroyed almost anything. At the time of this disruptive action, there was a lot of speculation about how a quick ans total switch to WP could possibly success. Now, two years later, WP7, WP7.5, WP8, and almost WP7.8 later, and 13 Lumia phones later, the ridiculous 2.5% market share very clearly show how big was the mistake. Not counting the manufacturers that have recently reduced if not throw away there WP product range.
This is in fact the real questions:
* How the Nokia board slowly changed to the point it elected a CEO that bring Nokia to his fall.
* How can the board allow disruptive and destructive action from the CEO without limitation or even a reaction ?
I wonder if Debian will not have the advantage to split the concept of release in different layers, like for example base, console-apps, desktop, graph-apps, etc... Upgrading the desktop layer should not affect too much the base and the graph-apps.
And the be competitive, Debian should probably propose more advanced revision of a package that the stick stable on to gave to the user more choice. I suspect that many enjoy a high stability but would like to install the last release of a small number of packages that bring important features to them. Yes there is backport, but it's a big mess to setup. The choice of the revision should be really simple from application like synaptic, apt-get, aptitude, etc...
Yes, long release cycle is not adapted to the today's fast moving projects. Stability is still a valuable goal, but there is not point in fixing too old revisions that users will not use and that upstream will nor car of anymore.
Aside of that, I found Debian still a very important project because it bring to the community a lot of very good and clean technologies. In this regard the multiarch will certainly be the beginning of a new area, especially now that a lot of different arch are coming to the mass in the form of almost anything but traditional PC.
The chance seem to be very small, but without that Wheezy will look like a old duck with his Gnome 3.4 when it will be released.
I hope that the Debian team will be cleaver enough to understand the advantage of providing a good classic desktop experience for people that will upgrade from Squeeze to Wheezy (I have tried Gnome 3.x and Unity and found them unproductive).
The idea they are a patent troll is completely and 100% absurd.
If you listen to a large amount of comments about Nokia this last two years, you will probably notice a dramatic increase of opinions that there is a lot of absurdities in the actual Nokia. The most absurd one is still that Nokia blindly ignore this reality...
Nokia is not a patent troll by any reasonable definition.
This was certainly true for a long part of the Nokia history. But the actual Nokia is something that have lost an extremely large amount of connections with the Nokia "mobile phone world leader" of the past. We are now forced to take notice that the actual Nokia is more and more close to the definition of patent troll. The latest new just confirm this trend.
This might not work if the kernel use a monotonic clock for the uptime like Linux do:
root@test:~# date -Is && uptime
2013-03-14T22:58:54+0100
22:58:54 up 5 min, 1 user, load average: 0.30, 0.89, 0.48
root@test:~# date 010100002000
Sat Jan 1 00:00:00 CET 2000
root@jtest:~# date -Is && uptime
2000-01-01T00:00:03+0100
00:00:03 up 5 min, 1 user, load average: 0.20, 0.82, 0.47
Uptime is still 5 minutes on this test virtual machine, not 13 years.
I appear that the reality is a bit more complex: http://en.wikipedia.org/wiki/High-level_waste
And if you take in account the risk perception by the population ? I think that you underestimate how much emotional can be the decision to live near a nuclear reactor with your children. And you cannot avoid emotional decision in a human community. The basis of this emotion is not only based on the number of dead, but also to the consequence on the survivors. In the two major civil nuclear accident, you have to add that the authority have massively lied about the scale of the problem. How can you expect that now people will believe pro nuclear saying that it's now under control ? There said that since the first civil nuclear plant and have been proved false at too many occasion.
Aircraft and cars for example kill a number of people each year, but even if the former is massively safer than the second, the risk perception is not necessary lower for the aircraft.
Sure that the amount of radioactive material released by Tchernobyl and Fukushima is far below the 4.2 million m3, but there will last far far longer and have been disseminate without bound in the environment. The consequence is that it's difficult to predict where there will be concentrated, mostly by organic processing, but not only. There is report that some new buildings contain irradiated material for example.
There are actually more like hazardous stacks of radioactive wreaks that need constant attention.
A truly decommissioned plant is safe and can be converted into an other uses.
Yes, there isn't any safe type of power generation, but the high concentration of long half live highly radioactive isotopes make the nuclear generation in a category of his own regarding long term risk.