They could fix the server problems in less than 24 hours if they started paying AWS more. But that would very, very quickly cost them more to run the servers than whatever money they have coming in. Breakdown from years ago (I think Bootcha and I discussed it) was something like $7m to $10m a month in server costs to keep a steady tick rate of 30. AWS CPU cycles are very, very expensive. Very. This would also solve a ton of physics and collision issues, since CryEngine ties a ton of that directly to server tick rate (stupidly, I might add).
Or they could install their own server hardware which would be a huge undertaking and cost them tens of millions in upfront costs.
Btw even lowballing at $5m a month in server costs, down from the $7m to $10m figure -- that's just for the playerbase they have now. "Server Meshing" or whatever would be...I don't even know. Prohibitively expensive, for sure.
Or they could install their own server hardware which would be a huge undertaking and cost them tens of millions in upfront costs.
Btw even lowballing at $5m a month in server costs, down from the $7m to $10m figure -- that's just for the playerbase they have now. "Server Meshing" or whatever would be...I don't even know. Prohibitively expensive, for sure.