The political aspect is another factor that I haven't even touched in my suggestions.
The top dog in terms of city sim with politics is IMO the Tropico series. Personally, I have only a few titles of the series and I got them in bundle packs, so I don't have almost any hours of gameplay accumulated in such a game mode.
See, this specific bug has never bitten me - juggling political power and approval ratings to be able to keep mayoring.
But I totally get the interest other people get in this kind of gameplay.
The problem is that's yet another layer of complexity and simulation to run on top of a game which is already limited on its performance envelope, because it can't go overboard its release-time baseline power requirements.
I think that even if we accept the current performance floor as a de facto immutable value for this game version, there are two concepts which should be explored in tandem, that may still be able to yield a higher performance ceiling to cram in more simulation.
1) Get a nerdy nerd cred which kicks ass on GPU Compute and try some massively parallel versions of the algorithms to solve game stuff which require to be done "a million times" per second. offloading e.g. traffic to the GPU might hit on the graphics part, but will also free up some PCU cycles, and the amount it eats on the GPU side will be less of an impact than the amount freed on the CPU side. It's worth at least exploring for the next iteration of the game.
1.1) If you hit the jackpot with the GPU Compute and can replace the traffic algorithm with a version which costs less compute-wise, allow the car limit to explode to 1 million on the street.
1.2) If you guys are already running e.g. traffic in GPU compute, I am barking at the wrong tree.
2) LOS , or Level Of Simulation - Just like there is this thing called "Level Of Detail" for graphics, there might be ways to creatively partition the game simulation to that a portion of it can run in a simplified model. If the city can be cut in "simulation districts" small enough so that they can be un/loaded as the user scrolls around, and the sim districts in "far view" run in a way that they cost 10% or less computing resources than a full blown sim district in "close view", and then much more can be simulated.
The top dog in terms of city sim with politics is IMO the Tropico series. Personally, I have only a few titles of the series and I got them in bundle packs, so I don't have almost any hours of gameplay accumulated in such a game mode.
See, this specific bug has never bitten me - juggling political power and approval ratings to be able to keep mayoring.
But I totally get the interest other people get in this kind of gameplay.
The problem is that's yet another layer of complexity and simulation to run on top of a game which is already limited on its performance envelope, because it can't go overboard its release-time baseline power requirements.
I think that even if we accept the current performance floor as a de facto immutable value for this game version, there are two concepts which should be explored in tandem, that may still be able to yield a higher performance ceiling to cram in more simulation.
1) Get a nerdy nerd cred which kicks ass on GPU Compute and try some massively parallel versions of the algorithms to solve game stuff which require to be done "a million times" per second. offloading e.g. traffic to the GPU might hit on the graphics part, but will also free up some PCU cycles, and the amount it eats on the GPU side will be less of an impact than the amount freed on the CPU side. It's worth at least exploring for the next iteration of the game.
1.1) If you hit the jackpot with the GPU Compute and can replace the traffic algorithm with a version which costs less compute-wise, allow the car limit to explode to 1 million on the street.
1.2) If you guys are already running e.g. traffic in GPU compute, I am barking at the wrong tree.
2) LOS , or Level Of Simulation - Just like there is this thing called "Level Of Detail" for graphics, there might be ways to creatively partition the game simulation to that a portion of it can run in a simplified model. If the city can be cut in "simulation districts" small enough so that they can be un/loaded as the user scrolls around, and the sim districts in "far view" run in a way that they cost 10% or less computing resources than a full blown sim district in "close view", and then much more can be simulated.
- 1