goldenarcher1 wrote: »Nope.
AD and EP were pop-locked,DC on 2 bars at prime time.
IC also has 'some' map updates which get send to the cliënts, but not near as much as those in Cyrodiil. IC zone flags, combat locations VS Cyrodiil bridges, keeps, scroll locations, combat locations, routes, camps, etc. All that has to go to EACH seperate cliënt, and is updated like every second. So when a certain Cyrodiil shard has many people on it, both combat metrics and zone metrics have to go over the same line for every individual player. That is a massive load of data going out to each player, every second.If Cyrodiil is still lagging without AOE's, while the rest of the game is not lagging with AOE's... there is only one thing left that could cause the Cyrodiil lag: The Cyrodiil map itself.
Maybe the Cyrodiil map updating constantly(keeps, routes, bridges, scrolls, combat locations, camps, etc), and sending that information to ALL cliënts constantly is what causes the lag. Kinda makes sense when you think about it, since ALL that extra information is send side-by-side with all regular combat information. While in normal zones/instances, only the regular combat information is send.
there was still ability lag in IC when large organized groups were running there so i don't think it's the map
Effects only need to be transmitted if inside visual range add margins and have client figure it out based on draw distance.
Crowd computing don't work here, it works for stuff who is not time critical but that is the main problem, even linking servers together probably create too much latency, now you can buy stronger servers but as they run on amazon cloud I'm not sure they can buy and install 100 core servers.
Raising the client in the room, most configure their graphic to look good overland out of combat. Pretty sure consoles are also set up much like this. Now we gotten lots of new motifs, outfit styles and mounts the last years. All have to go into video memory. And no its not only an hardware issue but lack of optimization here too.