@mainfold These days anecdotally it's more on par with 11 as devs get their implementations sorted, but yeah in the pc space there isn't much benefit most of the time. No real user facing advantage generally for sure. Might happen anyway because of the xbox one version but we'll see.
@SovietSpaghetti UE4 supports dx12 quite fully now so it'd be possible. That said it really doesn't make much difference outside a few very specific use cases. I wouldn't expect a noticeable difference in a game like insurgency. It's kinda like the dx10 of this generation imo haha. Not that useful to most of us, but laying the ground work for the next version.
@Nekrosmas Multi-threading in games is interesting, because games often require processing to happen in a fixed and predictable order so systems don't break. A really oversimplified example of this is if you were to simulate firing of weapons on a thread separately to the thread where you update the hitbox positions based on animation, at any given instant the code could be testing a bullet hit against outdated or future collision data depending on which thread was running faster or had less work to do. Realistically this probably wouldn't be a big issue but just using it as an example to describe how multi-threading isn't always practical in game development.
It's super good for stuff that doesn't need to be accurate like big crunching of numbers (a good place to use it might be spawning loot in a pubg style game because it's ok if that happens in the background while people play the game).
Another great example of where it's practical is Rainbow six siege. That game has a really cool system where when you plant c4 it calculates the destruction on separate thread in the background so it's ready to all explode when the timer counts down etc and doesn't hold up the main game thread.
So yeah multi-threading is great, but it's not a matter of just separating the existing 4 threads out onto however many a pc has but instead taking advantage of extra threads when it's possible (which is rarely every tick of a game)
Squad might be a closer example of tick rates, might be worth checking that out.
Good thinking. Had a quick look online and squad as of july last year drops to 25-40 at around 80 players, but they're constantly improving that game so wouldn't expect those to be final numbers
Source (last comment): http://forums.joinsquad.com/topic/27789-higher-server-tickrate/
Tl;dr a straight port of insurgency to ue4 shouldn't struggle to run at 60hz on the server if that's what the devs want to target.
Ue4 by default has an unlimited tick-rate only limited by performance/network performance/developer choice (which some of you probably already know). If Sandstorm keeps the same kind of scale as their other games I (anecdotally) can't see any engine based reason for 60hz to be difficult to achieve (source: I'm a dev, and have done most of the network replication programming on my own ue4 game.) If there was one factor that could affect tick rates on a server side it would be ballistics, depending on how authority is handled and what kind of rewind they use but it's only when dealing with fornite scale playercounts where getting adequate tick rates becomes extremely challenging.
Even epic working with their own engine have only achieved 20 hz in fortnite (source) but that's on a way more hectic level of complexity than what I'd expect from insurgency, considering they have to network all of the building/destruction systems as well as everything else. So I mean it's feasible that if the scope of Sandstorm is expanded far beyond the last game it could be more difficult to get a 60 tick rate but in my opinion it's unlikely.