Galactic Civilizations IV Galactic Civilizations IV
Alpha Access

Using More Cores To Process The Game

Published on Tuesday, November 9, 2021 By SSStreetFighter In Galactic Civilizations IV

I tried playing on a the second largest map against 16 opponents using a 2021 gaming laptop with 32GB of ram and the game got so slow after 200 colonies that it took 15 seconds for an order to show up in que after I clicked the button. I had a conversation with a computer science Reddit https://www.reddit.com/r/AskComputerScience/comments/qmwdcb/large_civilization_maps_from_a_computer_science/

During the lag my computer was only using 37% of the CPU. The reddit said that the problem might be that the game software is not optimized to use all of my Cores and that it is just burning out a few of them. My computer has 16 cores but the second largest map only uses 4. 

I tried playing to medium map and thankfully the menus registered at full speed but the next turn eventually got so slow that I clocked it at over 60 second per turn change and sometimes it would stop entirely and I would have to press escape then resume to get it moving again. So I had to abandon that game too. The medium map only uses 2 cores.

I doubt that a computer with less than 4 cores has been manufactured in the past decade. So I think the code should allow for separation into different subprocesses to utilize all the cores a computer has. The processes could be evenly divided in to 16 processes for people who have 16 cores. If the person has 8 cores then each could double up. If a person has 4 cores then 4 processes could run in each core. It just does not make since to use only 1/8th of my CPU on a medium map out of respect for the person with a 2010 computer who tries to run this game, and other problems will probably get in the way of a pc that old even if the you supposedly only need 2 cores. 

I am about to try the second smallest map. So far of the 15 maps I have started, I have only been able to complete 3 games and they were all on the smallest map.