GrahamKennedy wrote: But the limit is not simply in how fast the CPU can go. It's how fast the sensors can gather data, how fast the weapons can lock on, how long the weapons take to reach the target. If it was as simple as computers > people then there wouldn't be a place for manned fighters today. But in reality the effectiveness of defences isn't all that great.
Sensors have to gather data for the humans too.
The human has to wait for the weapons to lock on too (then approve the launch)
Manned platforms have to wait for the munitions to reach the target too.
Manned platforms have to deal with squishy slow organics. The primary role would be that humans don't trust machines to do the job, so they send the human there to make the final decision. Similar to officers and enlisted in the military. The enlisted is a specialist and will know more about a specific field, but the officer has the approval of the nation's government to make independent choices that affect overall operations. (I.e. the enlisted may know that the regular parts aren't available, so the vehicle has to be taken out of service. The officer knows the overall strategic situation, and lets the enlisted know they need the vehicle running somehow. The enlisted then has to figure out a way.)
So manned fighters would be out there for legal reasons, rather than practical. Plus also a heavy ECM environment would reduce communications, so you need a local control node, either manned or automated.
GrahamKennedy wrote: Sure, but again you're building a missile that's as expensive as a fighter. Your fighter expending munitions brings 90% of the cost back to be used again.
Is the expense due to needing to make sure the platform fly multiple missions, or due to the equipment? A fighter has to be able to fly multiple missions, reliably, and keep the pilot alive. A missile just has to fly once. Compare the F-18 ($29-57 million) with 6 Tomahawks ($3 million = $18 million). You need to send out an F-18 2-4 times to break even with a Tomahawk, but you also have to keep the support base set up, plus the maintenance, etc. A Tomahawk will sit inside a launch canister for several months without a problem. Pilots tend to complain when you do that to them.
At this point you'd have two types of missiles. The control platform that coordinates missile strikes, and has the expensive sensors. The expendable munitions that destroy the enemy. If the control platform gets shot down, you merely activate its backup and launch spares. If you lose the fighter you have to write a letter home.
Think more like an AWACS platform rather than direct combat.
GrahamKennedy wrote: You're also back to the issue of shield/armour vulnerability, as to whether the smaller ships are actually going to do you any good.
In which case you get into figuring out the smallest size platform that can damage the enemy, and deciding which of the following options you want:
1) reuseable/not (non-reusable means the more expensive sensors on it, the more it will cost to deploy, but the more likely you are to succeed. Cost-Benefit studies are needed.)
2) manned/not (Manned means you have a squishy slow processor than is legally allowed to operate independently. Legal and procedural studies are needed.)
(Hint - if selecting manned, you may want to go with reusable. Your pilots will appreciate it.)
Tyyr - those smaller corvettes would do a really good job with the enemy supply lines, plus being sent to dozens of systems to look for a nail that is sticking up. The ones that don't come back let High Command know where to send the heavy warships.