THANK YOU! Not only did we hit our original funding goal in 22 hours, but we made it through a full four stretch goals beyond that.
We'll continue providing updates through kickstarter -- of course -- but if you want up to the minute communications, please head over to our forums for the game.
Right at the end we hit one last stretch goal with the help of paypal backers: Interplanetary Weapons. (It's the only way to be sure.) That goes along with the awesome Nemesis, Spire, and music goals that were already hit.
You've all been amazing, and we look forward to a fantastic 2017 with you. :)
-- Chris and Keith
AI War II is a grand strategic RTS against an overwhelming, inhuman enemy who has conquered the galaxy. The enemy has made only a single error: underestimating you.
You must steal as much technology as you can, and take enough territory to fortify your bases and launch your attacks. But every conquest you make turns the attention of the AI ever more in your direction... so choose your targets with care.
We designed AI War II from the ground up for both solo and co-op play (up to at least 8 players), and put the emphasis on your smarts instead of your reflexes.
What Makes The AI Awesome?
The AI has a distinct advantage over you in a grand strategic sense, but when it comes to combat and tactics it plays by the same rules as you... and it's very devious.
The core conceit of the AI is that most of the time it's not aware of you. But pieces of its empire become aware of you at various times, and you have to then deal with those pieces in battles ranging from tiny skirmishes to thousands-strong epic encounters.
We've written in depth about our core approach to emergent AI, which keeps it from becoming predictable. However, on top of that we've layered a lot of specific tactical tricks to counter a wide array of player tactics.
The game has options to tailor the game to your skill level and playstyle:
Whether you're a masochistic ultra-hardcore grognard
Someone who enjoys RTS games.
Or just a space game fan!
Generally speaking, if you're winning all the time, the usual suggestion is to 'up the difficulty a notch.' Most players tend to enjoy the game best when victories are hard-fought or just barely lost. It's not about winning or losing, but about finding the right balance so that each campaign is an epic story of accomplishment.
Sidebar: The very highest difficulty is meant to be unwinnable -- and so top players enjoy their occasional victories at that level, then help us update the AI to prevent their prior tactics from working. We have a super fun time with that arms race, and it makes the AI at the lower difficulties more interesting, too.
Key Design Shifts From AI War Classic (The Original Game)
The Ark: your central command is now mobile, and can directly hack the AI. What could go wrong?
Making the AI's responses more fluid and aggressive.
Reworking armor and attack multipliers into new, intuitive defense mechanics.
Splitting Energy: globally-held Fuel drives your fleets, while per-planet Power allows for meaningful planet customization.
Reinventing Cloaking: the old long-term stealth ability becomes a dangerous assault mechanic.
First of all, we have the Design Document, which outlines what the launch build of the game is intended to be. This is useful even if you've never played AI War Classic. Notably, this explains the Key Design Shifts above in considerable detail.
For those considering the Personable AI reward (and for those who are simply curious) we have Appendix B - AI Traits, detailing the currently-planned traits that can be mixed and matched to create new AI personalities.
Key Technical Mumbo Jumbo
Lots and lots of multithreading to make awesome use of modern computers.
XML modding for game content, and advanced C# modding for several key logic zones.
DRM-free copies also available on launch, and day-one OSX and Linux in addition to Windows.
Robust scaling of the 3D graphical complexity so that it can look great while running well on older machines, and look even better on newer ones.
Already approved for Steam -- no Greenlight needed.
Thanks to fixing up a ton of usability issues, AI War 2 is well placed to deal with the first game's most common complaint: the killer learning curve. AI War is meant to be a hard but fair game, and the start menu should not be enough to defeat new players.
We’re coming at the problem from a few different angles. Usability changes will help smooth the difficulty curve once they’re in-game, as will quickstart options that allow you to jump into the game without fiddling with settings.
One thing we will not be doing is dumbing the game down. The point is to remove your need to fight the interface, so you can focus on fighting the AI.
We are pleased to announce that AI War 2 is built from the ground up to be open and easy to modify, while still maintaining multiplayer stability. Players can modify ships and structures, the design of AI personalities, and create tutorials, scenarios, and vignettes of their own.
You can write your own C# code (just as we do) to hook into various parts of the game, such as map-generation, which allows for much more robust modding than we had originally planned. Modders asked for this, and we listened!
The design document has a list of modifiable components that we identified before we even started making the game.
The project hit 100% funding in just 22 hours -- WOW! And our awesome backers carried us through a further FOUR stretch goals by the end. We have full writeups for the Spire, Nemesis, and Interplanetary Weapons if you want more details. We'll see what we can do about those solar systems post-campaign.
Bear in mind that the project funding total listed doesn't include the paypal contributions, which are what pushed us the last bit over the finish line.
Please note: rewards for backer tiers are as-shown in the sidebar. If you increase your pledge for the purpose of add-ons, that does not increase your baseline rewards. THAT said, if you pledge at least $260 through any combination of tiers and add-ons, we'll send you a copy of Stars Beyond Reach.
By popular request! Simply add an extra dollar amount to your pledge equal to the addon(s) you want, and then in the backer survey we'll ask you what exactly it is that you were adding on. Note that you can choose multiples of individual items (three cyber cyphers, etc) if you wish.
Extra Launch Copies
$20: One extra copy of the game at 1.0 release.
$35: Two extra copies of the game at 1.0 release.
$50: Three extra copies of the game at 1.0 release.
Add an extra $15 per extra such copy beyond the first 3.
Extra Early Access Copies
$30: One extra copy of the game at the Early Access release.
$58: Two extra copies of the game at the Early Access release.
$85: Three extra copies of the game at the Early Access release.
Add an extra $26 per extra such copy beyond the first 3.
Extra Alpha Copies
$48: One extra copy of the game at the private alpha release.
$94: Two extra copies of the game at the private alpha release.
$138: Three extra copies of the game at the private alpha release.
$180: Four extra copies of the game at the private alpha release.
Add an extra $40 per extra such copy beyond the first 4.
"Personable AI": If you've backed at a tier that is at least $160 (taunts), then add $40 to your pledge and you can note in the survey that you also want the "Personable AI" rewards in addition to whatever your chosen tier rewards are.
"Cyber Cypher": If you've backed at a tier that is at least $160 (taunts), then add $40 to your pledge and you can note in the survey that you also want the "Cyber Cypher" rewards in addition to whatever your chosen tier rewards are.
"Bronze Merc": If you've backed at a tier that is at least $100, then add $70 to your pledge and you can note in the survey that you also want the "Bronze Merc" rewards in addition to whatever your chosen tier rewards are.
"Silver Merc": If you've backed at a tier that is at least $250, then add $170 to your pledge and you can note in the survey that you also want the "Silver Merc" rewards in addition to whatever your chosen tier rewards are.
"Gold Merc": If you've backed at a tier that is at least $350, then add $270 to your pledge and you can note in the survey that you also want the "Gold Merc" rewards in addition to whatever your chosen tier rewards are.
Arcen (that’s Ar-KEN) Games started in 2009 with a lone programmer on a mission. Most real time strategy games at the time lacked artificial intelligence worthy of the name. Once you understood how the AI worked, the challenge, and most of the fun, disappeared. So Chris Park decided to make his own.
It would promote strategy over tactics, reward players for experimenting, and constantly surprise them. The game needed to be fun for Chris even if he knew the game backwards and forwards.
From its origins as a one man product, AI War Classic morphed into a community driven cult classic. Critical praise for the original is glowing, and it was one of the best-reviewed PC games on metacritic in the year of its launch.
Seven years, ten games, ten expansions to those games, and hundreds of patches to those games later Arcen Games is a proud independent game studio with fans all over the world.
Pablo Vega joined very early on and provided an awesome musical score for the game. He has since scored every other game we have put out.
Keith LaMothe is helming this new vision of AI War 2 as both lead designer and lead programmer. Fans of the original game know him well, as he basically was the force of nature behind AIW Classic from 2011 onwards. Chris may have started the AI revolution, but Keith gave it many of its most dangerous weapons.
Daniette "Blue" Mann has been with the company since 2013, and her graphic design talents are behind our increasingly-awesome interfaces in years past, as well as some of the coolest bits in our move to 3D.
Latest Updates from Our Project:
v0.500 Released: Major Milestone! Ship batch 7 of 7.
1 day ago
– Wed, Jun 28, 2017 at 01:08:20 AM
Chris here! Still not time for "beta" status, or Early Access, just yet. If you're waiting for the smooth gameplay experience, there's still a bit to go -- but that's now squarely the next thing on Keith's plate (and the GUI is a big part of that, yes).
There are two dire guardians in this one (which are two of the most nasty ones), and then there are four big honkin' Zenith golems. This includes the Dyson Sphere, the Zenith Trader... and yeah, the Devourer. :) Ho-lee smokes.
Bear in mind that these are the last of the ships until things are playing really well and the interface and whatnot have matured substantially. Sometime during the Early Access period when things are running well, Keith will work on those interplanetary weapons (a stretch goal), the spire stuff (another stretch goal), the Nemesis (yep: stretch goal), and whatever other minor stuff might be left that I've overlooked. I think basically everything else is in the game now, though.
Anyhow, so this is a very big milestone for us, because that means that Keith's focus can stop being on content, and can shift instead to refinement, usability, and all that sort of thing.
I'm finally done with all my performance-chasing stuff as of a little while ago, and as of last release the sound system is in place (though not with all sound effects yet, of course), so cumulatively this makes for quite a satisfactory 0.500, I think.
Post-Processing Visual Stack AGAIN!?
I just finished redoing that, right!? Well, it's been redone yet again thanks to some issues that are detailed in the release notes. The overall look is pretty similar to what it was before, except things are a little crisper and better antialiased, and there's a very slight vignette effect.
The big difference is in the bloom effect, which is generally more subtle now, but also more diffuse and thus a bit more dramatic in how it blends things around. Previously things were super intense in their bloom in a short range, but now they're intense in their immediate area, then fade out more gradually. Gives a more natural look.
Additionally, in order to deal with the flickering glimmering lights that could happen in some cases previously, the game now uses a much longer (about 3/4 of one second) temporal filter on the bloom, which can cause some really pleasing light trails that are purely a screen-space (and thus "free" in terms of performance) effect.
You can see that with:
The first shot in this post, which has the camera stationary and then the Ark moving away from it, leaving a bit of a ghosting trail from its engines. That's completely a screens-space effect.
The second shot in this post, which has the camera panning super fast to the right while the ships are stationary, and so you can see the glows being pulled to the left. This looks less good, but is also a lot less common of a case and generally happens so fast you can't see it unless you screenshot it while you're in motion. It does give a bit better sense of your own speed of movement, though, I'd say.
The third shot in this post (below) shows a stationary camera and both ships and shots that are moving, and how the glow trails help a bit with those. This is more subtle since the shots themselves are trail renderers in a lot of the cases there. It's noticeable in a still shot, but during gameplay I wasn't sure it was actually even happening.
If the temporal effect feels too strong to some folks, then I can easily make that a slider that folks can adjust in the settings, and we can debate what the best default setting would be if this is not it. For the moment -- and I'm too close to this at the moment to speak with any real confidence on the matter -- I like how this is looking, though.
Over the next half-week or so, Keith and I are going to be pretty quiet/absent, because of some other-life things that we need to take care of, respectively. But then after that he'll be working on the user experience and GUI and so forth, and I'm sure a bunch of mantis reports. I'll be working on mantis reports and more sound effects work, and getting the first voice acting in place, etc.
v0.450 Released: Sound effects at last! Plus video of it all.
6 days ago
– Thu, Jun 22, 2017 at 10:40:51 PM
Chris here! Just regular new version update this time, for a change of pace. :)
Release notes here. They are long, and in parts technical if you're interested in the innards of our new sound mixing system.
If you'd like a quicker look at the results of the sound mixing system, as well as a brief video-based overview of it, then here you go:
I had, naturally, hoped to get ALL of the shot graphics into the game today, and have their sounds linked up to them, and get the existing voice work integrated into the game... and of course that didn't happen. I was super tempted to delay the release because I didn't have every last thing in it, but already there's so much goodness in here that there was no point in waiting.
One of my favorite things was actually coded by community-member and top-bughunter BadgerBadger: the ability to name your savegames. We hadn't had time to get to it yet, so he said "heck, I'll do it myself." :) That was a side benefit to open sourcing those parts of our code that I hadn't really anticipated!
As always, more to come soon. The most complete source of info about the game's progress is a tossup between the dev blog and the forums (release notes aside).
"Early Access" Backers now have Steam Keys!
28 days ago
– Wed, May 31, 2017 at 11:19:17 PM
As of two days ago, actually. On the day of, the emails were sent out and I posted an update on the forums and in our comment chain here.
Chris here, by the way. :)
At any rate, I didn't have time to do a full post on that day, and there were some zero-day bugs with version 0.300 that led to us spending yesterday working on 0.301 and getting that out the door before I made this update. I have to confess, some of what I was working on with 0.301 were also visual updates, because I wanted to show off the new HDR visual stack.
That said, we're still not at the point of our original goal for starting actual Early Access on Steam, and we're at least a month or so out from that. As noted last time, we're giving out the keys to the EA-level backers from kickstarter and backerkit immediately, anyhow.
Gosh there's so much going on it gets hard to sum up. Those links have a lot of the really detailed notes if you want that kind of detail.
I just got my key. What's the state of the game?
Aha! A much easier question to answer, although you may be less enthusiastic at this one. The TLDR is that if you're looking to hop in and immediately play a complete game that needs polish, you're going to be unfortunately disappointed -- we're still in an extension of the alpha phase for the next month or so, and we'd love your help in giving feedback to help us get to the state described.
The sidebar on the left side of the screen, which should show your ship counts and so forth, is removed at the moment as we update it.
The interface in general is horrible, and you need to go into the PlayerData folder and edit your settings file manually if you want to make settings changes.
In very large battles, there are still some performance bottlenecks that we intend to address. All on the visuals/fluff side, not the core simulation, thankfully. We've gotten most of the heavy-hitters thus far, though.
Various ships are implemented but don't have real graphics, so they just show a little "rock" instead where the ship graphic would be. You can still see the ship icon and whatnot just fine.
All ship shot types use the same graphics right now.
No sound effects yet, and only a couple of music tracks.
Particle effects are also limited.
Ships currently just sit there in their squads, rather than flying around within their squads.
Not quite all of the baseline ships are implemented off our pre-Early-Access list, but we're pretty close!
There's no tutorial yet, and there's basically not much in the way of a lobby beyond map types and seed. You can't choose your player color yet, even. (Compared to the actual gameplay mechanics, we felt like these were ancillary things, but they are coming up in importance sooner than later.)
What Is The Purpose Of This Phase?
Short Answer: To make the game fun, which it is not yet. Please don't fret on that!
Long Answer: There's still a lot to do on the game obviously, as stated in the last blog post. But there's a lot of good stuff here to tinker with now, and we're really looking forward to having more people bashing on it. It's not a "fun, balanced game that just needs some polish" yet, but it will be really useful for us to have more people finding the pain points both in the interface (which is currently atrocious) and the actual gameplay flow (which, from a macro standpoint, is still pretty immature).
The underlying technology and components for making a fun game are here, and that'd a very critical step towards it being a fun and balanced game, but that's not where we are just yet. In a lot of respects we kind of reordered things: the underlying tech is somewhat more advanced and more polished than we had anticipated at this point, and that is pretty important because it gives us a better idea of what we CAN do in the engine. It gives us a better bounding-box for setting up things so that we can build an interface around what is possible, and have the scale of battles reflect what is possible performance-wise, and so forth. On that front, I'm super happy with where we are.
But yeah, the next step is to finish implementing the last of the "before early access" ships, and then to actually make a GUI that isn't eye-gouging as well as a game flow that doesn't have any obvious deficiencies. Right now there are some notable concerns about parts of the game flow regarding how you don't really need to keep territory as much as in the first game, and certain other bits of the feel from a strategic standpoint are "off." Some of that is just because xyz AI feature maybe isn't in place yet, but other pieces are more about the design of certain ships or mechanics. These are things we want to iron out before we go full-Early-Access, and we need the help of folks like you to do so. We're trying to streamline certain aspects of the first game, but we don't want to do that at the expense of what made the original game cool.
The engine for this one is so flexible that we could just recreate most of what the first game was if we felt like it, but we'd really rather not for a variety of reasons that should be apparent to anyone who tried to get into the first game and bounced off it, or who played the first game for a huge number of hours but wanted certain fundamental improvements. Now that all the basic frameworks are getting in place here, we're at a point where we can start thinking about those things.
New Tutorial: Creating Nebula Backgrounds for AI War 2
A how-to guide for modders or other players who want to create nebula backgrounds for either their own use, or to send in for official inclusion in the game. This requires version 0.302 or greater of the game.
Keith and I are both plugging away at usability issues, as is Blue when it comes to redesigning the interface. There are those performance issues that we're pursuing as well, and various other bug reports we're fixing up.
Basically the game isn't all that fun right now because you can't really work the controls all that well, and we want to get the interface out of your way. Thus far that has been serviceable-at-best simply so that we could get the underlying mechanics and gameplay and so on working.
Now most of the nuts and bolts are there, and a lot of the visual fluff (though not remotely all of it), and so it's time to buckle down on making it a fun game that is fluid to pick up.
By the way -- this might seem to be a backwards way to approach things, but given that this is based on another game that already was fun, but which had specific technical and aesthetic concerns we wanted to overcome, we started with the less-certain stuff. If this was a brand new concept for a game, no way would we have been doing things in this order! Always start with the parts you're least certain about, if you can.
When can I just sit down and play it for fun?
Not the bughunting-type fun? We'll let you know. We won't be going into Steam/Humble early access until it meets that criteria, so it shouldn't be too much longer. You can bet you'll hear about it from us when we hit that point, though.
AI War 2 Early Access: Good News and Bad News
about 1 month ago
– Thu, May 18, 2017 at 12:53:59 AM
Hi there! Both Chris and Keith coming at you, this time. So you know something’s up. ;)
What’s The TLDR?
1. We’re going to delay the actual Early Access launch on Steam until something like late June or early July. Previously it was intended to be May 29th.
2. BUT, for all the “Early Access” level backers from Kickstarter and BackerKit, we’re going to give you your keys on May 29th anyway, as promised.
Why The Delay?
Chris: The bottom line is that we don’t want to have a bad first impression on Steam. We’ve done that too many times in past years, to be frank, and we don’t have a solid reason to do that this time. So we decided not to. :)
Keith: The list of units that need to be implemented by the start of Early Access is dwindling rapidly, but not rapidly enough to leave a reasonable time to make the game generally fun to play before making the game generally available.
Aren’t You Changing Your Recent Story?
Chris: I’ve been saying, repeatedly with every update, that we seem to be on schedule. That things look good. Etc. But now we’re... delaying?
On the one hand, I think that my statements were pretty justified. In terms of actual feature implementation, we are on schedule and have been on schedule. Hooray, right?
On the other hand: something I failed to take into account was looking beyond the raw features themselves and into the deeper player experience. Is this game fun to play? How many people have been able to really beat on it in a playable state versus a prototype state? And other important questions like that.
Keith: When I started the first of the seven new-unit batches intended to get us ready for Early Access, I told Chris that we were on schedule. At the time we were because it looked like those batches were going to go very quickly. That's not what happened, but until I actually told Chris that my schedule evaluation had changed he didn't know. Translation: my bad, sorry.
What Benefits Come Out Of This?
Chris: Having the extra time is beneficial for a number of reasons, mainly in the testing and feedback areas relating to the user experience and gui. That’s more Keith’s area, so I’ll let him speak to that part.
From my end, we’ll be able to have that much more polish visually and aurally when it comes to the initial Early Access launch. Pablo is scheduled to complete a lot of music for the game in June, and that in itself is a big win when it comes to first impressions. We’ll have more of the visual whiz-bang done, and I’ll have more time to make a better trailer and do better screenshots.
This will also give Blue some dedicated time to focus on the GUI with Keith and players. This is one of the chief areas where we feel like we could really have suffered in early Early Access impressions otherwise.
Keith: This delay means we get to playtest the game much more thoroughly before Early Access. This obviously improves the initial impression that phase-launch will make, but also avoids that testing-and-iteration work getting delayed even further because we're putting out fires related to that launch.
We also get more time to work on a UI which doesn't cause Acute Ocular Hemorrhaging (because it won't have been put together by me)
Why Include “Early Access” Backers From Kickstarter Early?
Chris: Firstly, this is something that should help the delay be a little more palatable. We’re not trying to annoy anyone with this change in schedule.
Secondly, and much more importantly, we need a larger pool of testers. Our pre-Early-Access testers have done a great job helping us find a variety of problems, and BadgerBadger in particular has been a bughunting fiend. That said, since the game has been in a prototype phase rather than a true “let’s play the game end to end” phase, the enthusiasm for most people to put in a lot of hours is understandably low.
The May 29th date should represent the start of a period where you can play just to play… but there are going to be sharp edges all over the place that you cut yourself on. As a kickstarter backer who hopefully has an extra degree of investment in the concept of this game, you’ll tell us about these things so we can fix them using that extra month+ that we’re taking.
Then when we hit the actual Early Access launch on Steam, it’s a much smoother experience for the non-kickstarter/backerkit folks. That should be good all around, leading to more positive steam reviews and a better ability to accumulate even more feedback to help make things better prior to 1.0.
Keith: All that said, if you opted for an Early Access tier instead of an Alpha tier because you didn't want to play the game until it was at least somewhat polished, it's better to wait to play until later. Having the key early won’t mean you can’t do that; we’ll keep you apprised of development.
What’s The Downside?
Chris: From an external perspective, if someone didn’t back the kickstarter, and doesn’t want to via backerkit, they’re now delayed in being able to play the game. From an Arcen standpoint, we’re now delayed in gaining income from Steam and Humble. I’d rather have the long-term goodwill rather than short-term money, but nonetheless it is uncomfortable. Overall, since we’re able to give the Early Access tier backers their keys on the intended date, I think the downsides are blessedly few.
Keith: It means we're not doing what we said we would do, which is never good. That said, we hope you understand why this is the less-bad of the available options.
What Else Comes Out Of This?
Chris: This isn’t truly related to the delay, although the timing is nice. But we had some players on the forums asking about having more voice acting for things like units accepting orders. I’ve started the process of getting those lines written (by me) and performed (by a dozen or so voice actors), and we should now actually have time to get all of those lines in place prior to Early Access. I think it’s something that adds a lot of personality to the game, even though I was resistant to the idea at first. And yes, you can turn them off if you don’t like them. :)
I also just want to add that Keith continues to amaze me with some of the revisions he has been making to various mechanics from AI War Classic. The way that warheads are now handled is the most recent example. These things are super exciting, but since they are new I want to make sure that folks have a chance to actually play with them before we hit that first “public Early Access on Steam” period.
How Does This Affect BackerKit?
We have the ability for anyone to preorder the game via backerkit, for that interim period between the kickstarter ending (back in December) and the game becoming available on Steam Early Access. This also allows existing or new backers to select add-ons for extra bonuses for themselves or extra support for us, or both. For those who need to use paypal, we have that too.
The plan on both counts was to disable those on May 29th, when the game hit public Early Access. Since the public Early Access date is being pushed back, we’re going to push the closure date for those two services back as well. We’ll give you advance notice, but basically when the game appears on Steam and Humble, it will disappear off BackerKit and our own paypal link.
Why? Mainly because we can’t indefinitely keep adding user-requested content of the sort that comes with those add-ons. We’re fine for now on that front, but it can’t be an open-ended thing, and the start of the traditional storefronts is a natural closing-off point for this more funding-raising style options.
Chris: I do want to apologize for the disruption, even though as kickstarter backers you’re probably used to projects being delayed. It hardly matters what you’re used to; we still try not to do it to you.
At the present time I don’t see this having any negative impact on our plans for a 1.0 launch in October, but I’ve been wrong before, clearly. If something does cause a delay past October for 1.0, we’ll do the same thing as now and get the 1.0 backers their keys on the original timetable rather than whatever the final market timetable morphs into. It’s very much in our best interest to keep to the October schedule if possible, financially-speaking, but we’re also not going to do anything long-term-stupid for short-term gains, if we can help it.
Keith: What he said. And thank y'all very much for your support. Rest assured, the Player Extermination Apparatus will be happy to solve all your "my other games are too easy" problems when the time comes.
Chris: Once again, on behalf of the entire Arcen team, I’d just like to thank everyone who has supported us, and who continues to support us. We wouldn’t be where we are without you, and we’re super excited about how this game is turning out so far.
Taking a look at some AI War 2 ships during late alpha.
about 2 months ago
– Fri, May 12, 2017 at 07:53:48 PM
Chris here! This is just a video looking at a variety of the ships in AI War 2, or at least the graphics for them. These are in the version 0.124, which will come out early next week. It's presently late alpha for the game (in the pre-Early-Access sense), and so these are coming up to a much more polished status now.
As part of our testing thus far, one thing that we've discovered is the need to use GPU Instancing. That was something that I hadn't been sure if we'd need or not, and I've mentioned it since our first kickstarter for the game. I wanted to try to get away with dynamic batching, which is compatible with OpenGL 3.x and DirectX 9 and DirectX 10. However, the performance just wasn't good enough, even in battles with only something like 5000 ships versus maybe 2000.
A few passing bugs aside, the performance was still better than AI War Classic with that scale of battle on the simulation side in particular, but GPU instancing became a clear need. So now the game is going to use that, which requires DirectX 11 or OpenGL 4.1, and basically hardware from 2010 or 2011, depending on your exact hardware and OS.
Realistically you needed hardware from that era at the oldest anyway in order to handle the CPU processing, so this really should be a moot point, but it was a bridge I hadn't wanted to cross unless it really became clear it was needed. Well -- now it's clear. :)
A bug in the GUI sidebar aside, I was getting about 30fps in the aforementioned battle using dynamic batching. This is on a latest-gen i7 with a GTX 1070. Now with most of the stuff working with GPU Instancing, I get around 80 fps. There are still thousands of wasted draw calls because of some of how I'm handling my custom sprite system at the moment, and I expect to get my machine running that same scene at 120 or 140 fps by sometime next week. Knock on wood. :) But it definitely seems like that will be what happens on my rig, based on all my tests thus far.
Anyway, so we get to the question of how big battles will be able to be, and to that I still have the answer: I really don't know. For a variety of reasons, we can do larger battles than AI War Classic if you're running them on modern machines. On a machine past a certain age (maybe from 2012 or before?), then the battles of Classic might be larger in terms of what your machine can handle. I'm not sure. The newer your machine gets, though, and that's looking to the future as well, AI War 2 starts pulling further and further ahead. This switching to GPU Instancing is a huge amount of future-proofing in and of itself.
Overall we just have a ton of performance optimizations and multithreading in the game already, and it's built around a variety of design concepts that lend themselves to larger battles than the original. We still do hit the occasional hiccup, like the sidebar thing, though, which makes performance absolutely grind to a halt for a bit. That's one reason why we do the alpha, though; so we can fix things like that, and they never last long. :)
All in all, we're looking good! I'm excited about the recent changes, even if I am apprehensive about any potential backlash by someone angry about the system requirements change.