With the announcement made (last year I think) that oracle would charge for each processor in a multi-core processor, combined with Sun's push towards CMP, Sun could be trying to make their servers more attractive towards potential DB owners...
Most auto manufacturers are EXTREMELY anal about the reliability of their parts - I believe almost any semiconductor that goes into a car will run single digit DPPM - that's ~2 orders of magnitude better than most chips that go into $30k+ servers, and another 2 orders of magnitude better than chips that go into consumer workstations/pcs.
Even with the fact that there are many chips in cars, there should be very few that suffer from an electrical failure in any of the controllers.
Dual core US IVs have been out since last Feb - second generation come out this month, and third generation this summer.
IBM has had dual core chips out for about the smae amount of time, I think...
Power goes to two primary areas
Leakage will account for ~1/2 of the power loss - at sub 90nm, transistors do not look much like a switch - they look much more like a resistor with 2 different possible states - this power will be roughly constant
The other half of the power goes to charging/discharging all the internal capacitances - setting a voltage on the gate of an individual xtor is just placing a small amount of charge on a small capacitor, and any time a gate changes states, you have to expend some energy to make this change 200M xtors X 4GHz = lots of state changes - this power will increase linearly with operating frequency
since most photoelectrics work off the same principle (photon moves electron to a higher energy state, and so forth) fluourescents would probably work fine.
The difference with IR is that their wavelength is much lower - thus the energy is lower, and absorption is that much more difficult (long, detailed explanation omitted)
SOC work has already reached this point for many electronics devices - cell phones are about the highest level things that I can think of that can currently be built from a single IC
This is also being seen in the processor market, especially on the server side, where chips can be significantly more expensive, with things like memory controllers, network controllers, etc being put on-die on a lot of next generation processors
The main problem is that combining all these parts into a single chip raises the cost out of the range that the typical consumer would consider. Since every new process node provides an ~50% reduction in die size, and assuming defect densities are fairly constant, it is conceivable that prices would be low enough to make it to the mainstream market sometime late in the 45nm or early in the 35nm generation (~3-5 years)
Slashdot Mirror
With the announcement made (last year I think) that oracle would charge for each processor in a multi-core processor, combined with Sun's push towards CMP, Sun could be trying to make their servers more attractive towards potential DB owners...
Most auto manufacturers are EXTREMELY anal about the reliability of their parts - I believe almost any semiconductor that goes into a car will run single digit DPPM - that's ~2 orders of magnitude better than most chips that go into $30k+ servers, and another 2 orders of magnitude better than chips that go into consumer workstations/pcs. Even with the fact that there are many chips in cars, there should be very few that suffer from an electrical failure in any of the controllers.
Dual core US IVs have been out since last Feb - second generation come out this month, and third generation this summer. IBM has had dual core chips out for about the smae amount of time, I think...
Power goes to two primary areas Leakage will account for ~1/2 of the power loss - at sub 90nm, transistors do not look much like a switch - they look much more like a resistor with 2 different possible states - this power will be roughly constant The other half of the power goes to charging/discharging all the internal capacitances - setting a voltage on the gate of an individual xtor is just placing a small amount of charge on a small capacitor, and any time a gate changes states, you have to expend some energy to make this change 200M xtors X 4GHz = lots of state changes - this power will increase linearly with operating frequency
since most photoelectrics work off the same principle (photon moves electron to a higher energy state, and so forth) fluourescents would probably work fine. The difference with IR is that their wavelength is much lower - thus the energy is lower, and absorption is that much more difficult (long, detailed explanation omitted)
SOC work has already reached this point for many electronics devices - cell phones are about the highest level things that I can think of that can currently be built from a single IC This is also being seen in the processor market, especially on the server side, where chips can be significantly more expensive, with things like memory controllers, network controllers, etc being put on-die on a lot of next generation processors The main problem is that combining all these parts into a single chip raises the cost out of the range that the typical consumer would consider. Since every new process node provides an ~50% reduction in die size, and assuming defect densities are fairly constant, it is conceivable that prices would be low enough to make it to the mainstream market sometime late in the 45nm or early in the 35nm generation (~3-5 years)