Ashes of the Singularity: New RTS from Stardock and Oxide Games


Ashes of the Singularity was announced today and is being shown at GDC this week. It's using Oxide Games' Nitrous engine which will allow thousands on units to battle onscreen.

Ashes of the Singularity is a new real-time strategy game in which players fight for control of entire worlds. Multiple armies battle it out across vast landscapes as the player tries to take control of the key regions of the map to obtain their valuable resources.

Thanks to Oxide Games’ Nitrous engine, which allows for thousands of units to clash onscreen at once, each acting independently with multiple individually targeted weapons systems and ballistics models, Ashes of the Singularity is able to deliver a game experience that is both deep and approachable.

Players are able to manage their massive armies through a new game concept called the "meta-unit." The meta-unit allows Ashes’ enormous scale to translate into fun gameplay. By grouping units together into a meta-unit, you make them aware of each other and alter their behavior to fight intelligently together and support one another. Because a meta-unit can take care of itself within its means, your attention is freed to direct the overall war effort – a good thing, as Ashes of the Singularity’s maps and unit counts are an order of magnitude larger than in a traditional RTS.

Servo: Sci-fi RTS from Age of Empires creators

Whoah. You guys are cranking out the strategy games.


pew pew!


This looks awesome. I really liked the StarSwarm demo, so it’s cool to see it getting used in a game finally.


I wonder if they took some devs from the SupCom series…


So much RTS goodness.


The teams are really impressive.


Glad to see Stardock on the offensive all over the place. Good luck and Vae Victis!



The only bummer on this one is the 4 player limit, but other than that it looks great.


Indeed, I immediately thought (hopefully? wistfully?) of ‘[B]Supcom 3![/B]’ when I saw that screenshot. And that’s doubly true after reading the press release about thousands of units, independently targeting weapons and a ballistics model.

The “meta unit” language seems a bit awkward to me (since it basically sounds like a [I]smart [/I]unit formation, so they could’ve used a military term like division, company, squad, battalion, or brigade instead), but nit-picking aside, the concept sounds great.

I had mixed feelings about some of the ‘smart’ unit behaviors/targeting/mechanics in Sins of a Solar Empire (especially the counter intuitive shield system that discouraged focus fire), but it had the distinct advantage of letting the player focus more on the macro strategy of fleet composition and deployments (and microing capital ships of course) instead of fiddling with each and every little frigate.

Rise of Nations also stands out in my mind for its smart unit formations that could handle themselves pretty well without tons of individual unit micro, so any RTS trend in that direction is good news IMO.


Behind the scorching beauty of Oxide’s Ashes of the Singularity. :)


That does look pretty, however I wonder about the mechanics and UI. In the video they said they wanted you to not manage battles, but wars. Planetary Annihilation biggest fault was the UI was not even remotely close to handling such a task. The other thing I would be concerned about is if there are any new mechanics. On the surface, it would just seem to be a very generic RTS that simply has 1000s of units.


Yeah, it’s a very early look and there’s still lots to work out in regards to UI and such. For the announcement at GDC, it was good to show the scale and capabilities of using the Nitrous engine and what can be accomplished. I can say a generic RTS it something it won’t be, and hopefully we’ll have some more details to show off real soon. :)



can they beat the unit design of ta/supcom?


I guess we’ll see? I don’t see how it needs to, though. Given how little we know beyond “huge RTS”, I can’t stay that’s on my to 5 list of things to wonder about.


Hehe. I’m glad someone else posted a SupCom video…
I was about to do the same last night because I found some of the hardware/tech claims in that AMD promo video a little strange*. However, while searching for a suitably impressive SupCom video, I got distracted watching a 60+ minute cast of a 4v4 SupCom FAF game on Seton’s Clutch. Now I just want to get back into SupCom.

*(They kept referring to “thousands of units on screen” when they were only showing 100-200 at most. Maybe they were counting projectiles? Regardless, games like SupCom, Total War, and even Sins of a Solar Empire have already put hundreds or thousands of units on the screen. I also noticed that all of the ‘land’ units seemed to be hovering, which seems like it might be a concession to simplify unit animation and makes it further reminiscent of SoaSE).

But I don’t want to be too snarky or negative. I’m glad they’re sharing footage, and I’m happy to see [B]any[/B] company making [B]any [/B]moves towards massive-scale RTS. From my perspective, there’s still only one good game in that niche (Supreme Commander FA), so I’d be happy to see a fresh take… And even if the map size and unit numbers don’t increase, it’d be great if another SupCom-alike game could be released with an updated engine that uses multi-core CPUs and >2gb of RAM.


There’s nothing anyone can do to Supreme Commander and FA to keep it from slowing down, as it’s a limitation of its architecture because the majority of its AI, targeting, UI, and general logic was written in LUA. That’s the bottleneck. All of that was moved into C++ in SupCom 2, which is why you can have 8 players with 4000 total units with little or no slowdown. (And despite what people seem to believe, Seton’s Clutch is the same size in both games.)

People have hacked SupCom 2 to have higher unit caps, and it handles it fine.

Here’s 1000 unit cap, 8 players:

Adding more players in multiplayer is a networking issue, not a performance one. Each player adds an exponential amount of network traffic, as the peer-to-peer networking sends every player’s data to every player. You’d need to switch to client/server to fix that, but then you’d have the issue of server loads and lag, units warping around, etc. And there aren’t a lot of games tracking 8000 players simultaneously; games like Planetside 2 aren’t necessarily that interested in people outside of a certain radius, whereas an RTS requires precise knowledge of everything on the map at all times (including projectiles, assuming your game is physically simulated like SupCom is instead of deterministic like StarCraft and Age of Empires).

And yeah, making everything a hover unit simplifies animation and pathing, which get crazy expensive as you increase the unit counts. But anyway, there’s nothing on the technical side that should keep this game (or any other) from improving the visual quality over SupCom FA or SupCom 2.


@Steve: Thanks for the explanation. I knew that SupCom2 could handle tons of units with good performance, but I’d never heard that a lot of the performance gains were from switching to C++ instead of LUA.

Anyway, I’m not even particularly concerned with a graphics upgrade (SupComFA still looks pretty good IMO), but watching that 4v4 FAF cast last night reminded me of the ‘sim’ speed slowdowns that FA suffers from when games get really huge (the video was ~90 minutes real time to watch ~60 minutes of game time). SoaSE suffers from similar constraints when games get really big, so that’s the one thing I’d hope to see resolved with this fancy new hardware/APIs they’re touting.


The numbers of units to expect in an Ashes game is over an order of magnitude higher than than what’s been previously done. Most of this has to do with the parallel rendering. So, for instance, on a large map, we can handle 20,000 + individual units (not counting projectiles or the independently tracking weapon systems on each unit that can have their own firing solution).

My love of TA and SupCom (I was in PGL for TA) are well known so this is no negative on those games. But from a scale perspective, this is something of a different kind. I talked to Chris at GDC and groveled for a TA 2. The main thing here Ashes has the benefit of having been designed for multi-core processors and 64-bit from the start.

And that’s not counting the fact that the terrain is procedurally generated (down to the erosion), that it’ll come with a map editor, that each shot is its own light source, that the units have true line of sight (as opposed to radial so the actual shape and level of the terrain affects what you can do, the firing solution of the units will filter out units they can’t hit (i.e. units won’t attack sides of mountains), and on and on.


Also, bunch of Q&A today over in the Ashes forums today: