Mainframe Computer

I'm not really happy with the 'clusters of computers' rule in the Central Supply Catalog. In High Guard, ships computers take no tonnage and allow a Core/100 computer at TL 15; but this is impossible with the rules as written in CSC.

A TL 10 'Superior' computer/4 (the very best listed 'mid-sized' computer) masses 5 kg; figure about 1.6 liters -- that gives 8236 of them per dTon. They start at computer/4 and 13 doublings (2^13) is 8192 computers combined -- for a grand total of Bandwidth/17.

If we allow building clusters out of 'portable computers' then a TL 14 Computer/5 masses 0.5 kg. This is ten times better, but only gets us to 2^16 computers in a dTon -- maximum of Bandwidth/21.
 
I'm not really happy with the 'clusters of computers' rule in the Central Supply Catalog. In High Guard, ships computers take no tonnage and allow a Core/100 computer at TL 15; but this is impossible with the rules as written in CSC.

A TL 10 'Superior' computer/4 (the very best listed 'mid-sized' computer) masses 5 kg; figure about 1.6 liters -- that gives 8236 of them per dTon. They start at computer/4 and 13 doublings (2^13) is 8192 computers combined -- for a grand total of Bandwidth/17.

If we allow building clusters out of 'portable computers' then a TL 14 Computer/5 masses 0.5 kg. This is ten times better, but only gets us to 2^16 computers in a dTon -- maximum of Bandwidth/21.
That's why I stopped at TL10... well, that and not making the table too long. Buy a ship's computer instead. Assume they're built on a different architecture ( I refuse to call it positronic). I just wanted to get some scaling for non-ship stuff that wouldn't obsolete the ship stuff and early stellar levels.
 
That's why I stopped at TL10... well, that and not making the table too long. Buy a ship's computer instead. Assume they're built on a different architecture ( I refuse to call it positronic). I just wanted to get some scaling for non-ship stuff that wouldn't obsolete the ship stuff and early stellar levels.
Yeah. Not an easy thing to fix without messing something else up.
 
Sadly never, not even up to the 57th century...

Not saying there is never going to be a way to get FTL travel or comms, but not within the framework of current quantum stuff, so I cry inside every time it is suggested.
Current quantum stuff, no, but the rate at which humans are discovering weird new quantum things is nothing short of astounding.
 
How about we have a Morse code made up of alive and dead cats, to communicate instantly across the cosmos?
The problem is that even if your cat is alive, all you know is that the other cat is dead, but that doesn't convey any useful information, because at the far end of the cat killing contraption, finding out that your cat died can happen at any time. And if you watch a bunch of cats keel over in a row glass boxes, then the whole uncertainty thing collapses on you - and half the cats on your end are dead. The rest are pissed.
 
I think computing power for a starship describes what it is capable of processing-wise. Storage is virtually unlimited (though practically that's not reasonable, but let's go with it).

Computers could easily be distributed via nodes so that damage to any one of them reduces rather than destroys it. At least that's how the game is structured. Who knows if technically in the future it makes sense to recreate "mainframes" due to some advantage. In Star Trek:TNG the computer was massive and occupied decks worth of space - at least according to deckplans. CT postulated 1979s tech, and today we've changed that up to racks and racks of servers, but each one can be physically independent.

One can argue either way really for this. Unless they crack transmitting data IA some method other than wires and speed of light, placing your components adjacent or very close to each other does offer an advantage. But the distance on a ship would be small, and that advantage may not be of much value.
 
"Portable computers" is made to be small and light. Cost of this is the cap on the highest tech that would work within the chassis.

HG says ship's computers are part of the furniture. Bit like a fitted kitchen or fitted wardrobes. Because size and convenience is not such an issue, then they can house smarter tech and higher bandwidth. RAW: compared to the tonnage of a ship's hull, computer weight is negligible, or practically near zero.
 
But the distance on a ship would be small, and that advantage may not be of much value.
Distances that matter in electronics are more like a half of the circuit board. RS-232/488 might be across a room. Ethernet - across a building. Internet - across countries. Across nebula - X-boats AFAIK.
 
"Portable computers" is made to be small and light. Cost of this is the cap on the highest tech that would work within the chassis.

HG says ship's computers are part of the furniture. Bit like a fitted kitchen or fitted wardrobes. Because size and convenience is not such an issue, then they can house smarter tech and higher bandwidth. RAW: compared to the tonnage of a ship's hull, computer weight is negligible, or practically near zero.
Right, so if I am designing a Military Base, how can I determine a Military Base worth of Bandwidth?
 
Distances that matter in electronics are more like a half of the circuit board. RS-232/488 might be across a room. Ethernet - across a building. Internet - across countries. Across nebula - X-boats AFAIK.
There are exceptions to that. A prime example is in stock trading. There are traders who have moved their servers next door to the NYSE to take advantage of transmission lag from say Chicago or other locations. It's slight, but measurable (and profitable). Even with fiber optic cables and fiber optic switches in the network.
 
"Portable computers" is made to be small and light. Cost of this is the cap on the highest tech that would work within the chassis.

HG says ship's computers are part of the furniture. Bit like a fitted kitchen or fitted wardrobes. Because size and convenience is not such an issue, then they can house smarter tech and higher bandwidth. RAW: compared to the tonnage of a ship's hull, computer weight is negligible, or practically near zero.
If you were addressing my examples from CSC, I think you missed. The point was that even 'Takes up a full dTon' (something that ship's computers explicitly no longer do in Mongoose) there is no way to build a Bandwidth/100 computer. I illustrated this with max-tech 'Mid sized' computers (TL 10, which should be able to build a Core/50 computer for 60 MCr) fails, at Bandwidth/17 and 81.9 MCr.

Bending the rules a little to allow using more compact and sophisticated 'Portable' computers also fails. Using max-tech portable computers (TL 14) fails too; TL 14 should be able to produce Core/90 computers for no dTons and 120 MCr -- instead, a dTon of clustered portable computer/4s will only get to bandwidth/21 at a cost of 327.7 MCr.
 
Last edited:
There are exceptions to that. A prime example is in stock trading. There are traders who have moved their servers next door to the NYSE to take advantage of transmission lag from say Chicago or other locations. It's slight, but measurable (and profitable). Even with fiber optic cables and fiber optic switches in the network.
Yes, it is only an approximate granularity. After all, how big is a hall compared to a room? Or how big is plaza compared to a building? For example.

Moving transmitters and receivers around to improve peak performance has been happening all through communications history. What do peopel do when their WiFi or Blutooth goes down?
Ans: Fist step, they move their device closer.
 
Right, so if I am designing a Military Base, how can I determine a Military Base worth of Bandwidth?
The advantage of the current base rules is that you can make a building a separate entity with its own computer, thereby giving it all the bandwidth of the chosen computer. Different computers for sensors, weapons, screens and coordination/flight control.
Yet still allowing them to communicate with each other.
 
The advantage of the current base rules is that you can make a building a separate entity with its own computer, thereby giving it all the bandwidth of the chosen computer. Different computers for sensors, weapons, screens and coordination/flight control.
Yet still allowing them to communicate with each other.
I don't really see that; High Guard covers (very large!) ships and space stations -- and the rules are clear on only one main computer, and one back-up computer. Vehicles doesn't address computers at all. CSC comes up short when trying to build large super-computer clusters.

Maybe I am just missing the rules you are looking at; w\here are you finding rules for buildings, and letting their computers work together?
 
I don't really see that; High Guard covers (very large!) ships and space stations -- and the rules are clear on only one main computer, and one back-up computer. Vehicles doesn't address computers at all. CSC comes up short when trying to build large super-computer clusters.

Maybe I am just missing the rules you are looking at; w\here are you finding rules for buildings, and letting their computers work together?
Land bases. You can divide them up as you like. Drinaxian Companion.
 
Land bases. You can divide them up as you like. Drinaxian Companion.
A computer in every hut!

I'm planning on some better integrated hut building rules as part of the Vehicle Handbook update. But you will still be able to put a computer in every hut. And if you install a galley (er... the cooking type, not the type with oars), you can put a pot with a chicken in it in every hut too.
 
How about:
Clusters of Computers: A cluster of computers is made up of a number of computers of identical bandwidth; dissimilar computers can participate, but are limited to the bandwidth of the least capable computer making up the cluster. Clusters start with the bandwidth of the least capable computer as their 'base value'; the bandwidth of the cluster increases by the full base value every time the number of participating computers are doubled. It is standard practice to automatically drop individual computers from the cluster if their bandwidth drops to the point that they would hinder the cluster.

It doesn't really fix the underlaying 'cannot build a Core/100 at TL 15' issue, but it comes closer.
 
Back
Top