The DOD was unable to procure more because Intel shut down the line in 2004.
Before the plant was closed down, given the realities at the time, the US would have procured as many chips as it thought necessary to sustain the fleet planned at the time.
They bought the last 820 i960MX chips off the line prior to the line shutting down, which was enough to produce only 155 F-22's.
This does not add up. At the time in 2004 the forecast for builds was higher. The decision to shut down the F-22 production line was taken after this time and the build capped at about 195 total aircraft (8 test and 187 in service). 185 aircraft are flying today...so only getting enough for 155 aircraft just does not add up.
The DOD can't buy any more i960MX chips because the factory that produces the chip no longer exists. The DOD planned to fund a replacement architecture, possibly based on the F-35, but that died quietly sometime after 2004. Thus any future production of the F-22 will require a new avionics architecture to be designed from the ground up for it. When the DOD capped production numbers, all momentum for a potential avionics upgrade was lost because of the very small fleet size.
And yet, as I said, the Cap happened after the dates you indicated and the number was higher than the 155...yet they are flying.
The three options are basically funding a massive avionics replacement program for the entire fleet, continue paying significant sums on new iterations that have relatively minor increases in capability, or canning the fighter altogether. I know the last one is unthinkable now, but I wonder how the AF might feel in 2020 when its getting fully functional F-35s and faces another costly upgrade to the F-22's avionics architecture.
As I said, when the military procures a system like this, they look out over the life of the product and procure enough of the critical parts needed to sustain it through its life.
So, making increments and changes beyond what was procured when they closed the thing down can be difficult, but maintaining what they had should be doable.
Also, although the factory for that chip is gone, given the amounts of money they are spending, if they were serious about the need, those chips, or something to replace them could be made. But, they can also come up with other solutions as you suggest. Making new chips, either the ones needed or new ones to replace them, for the fleet is probably going to be too expensive politically unless a decision was taken to build another 100-200 aircraft.
The US could definitely do what was necessary for that if we had the political will to do so. But, sadly, that is what is lacking and so the program will move forward, probably with very expensive, incremental improvements.
In fact, the current plan for upgrades is to go to 3.2A in 2014 which will focus on electronic warfare, communications and identification. Increment 3.2B is planned for 2017. Increment 3.2C (which I believe has now been changed to 3.3) is planned for 2019. That 3.3 update in fact targets starting the migration of avionics to an open platform, allowing features to be added at later dates. I believe the current budget to get to 3.3 is something like 6.9 billion. 3.3 is now slated to be bid out.
