PA Benchmarks - How each hardware performs

I was looking for any benchmark of PA on internetz, but I can't find even ONE. So, why not make one, or at least topic on forum where each person tells its average fps on specific games (1vs1/big FFA long into game, etc.). Tests with recording video or symilar would be the best, but anything is useful.

Personally I'm interested how R9 290x and i5-4590 performs in the game. Or any i5 and R9 290x. Im upgrading soon, and I would love to see some benchmarks.

I found difficult to benchmark PA, as there can be as much as 20% variation in fps in identical games.
Also whether you plan to play on the internet, play on a local server or host a local server (or all at the same time) can change what hardware you need.

The main thing I can think of about PA is that the amount of RAM you need can be very high, and that PA is one of the few games i know where memory bandwidth is a bottleneck.

Now save games / offline replays are a thing we can actually attempt A benchmark... Someone should post a replay file and specify a time. Then everyone opens the game, and posts a screen grab of fps at the specified point

Edit we should do this at a set of resolutions and settings too so we can compare entry level kit like laptops as well as full on desktop rigs.

Yeah that would make sense. Actually I'd like that.
I have a couple questions though :
- Would a replay of a game be as hard to run as the original game? I heard that everything was regenerated locally on the fly during replays, but most games are played on playfab servers.
- Whats the best way to share a replay?
- Does anybody know a good software for graphing my fps? I use RivaTuner's OSD, but it cant graph afaik.

I think for the client it makes no difference, for the server it is a difference of night and day. Playing a replay is like playing a ready made set of simulation results, so cannot be compared at all.
PA isn't easy to benchmark indeed.

MSI Afterburner? But I'm not sure it fits your needs perfectly.

Saves and replays are a good idea, but anything will be useful for sure. Even if it varies. This game is very random, and 400 Dox on one screen is not uncommon.

Actually Rivatuner OSD is the OSD that MSI afterburner uses. I didnt knov it could save the fps. How do you get the file then?

I was thinking from a client side benchmark.. agreed server is a different problem altogether...

How about how other RTSs? Starcraft probably has a rather unified fps, considering how there is a unit cap, but for games like supcom and TA, which didn't have a unit cap (did it? Im confused. It seems like it doesn't), how easy was it to benchmark the games?

Supcom and TA had unit caps?

Edited for clarification

No no I mean- Supcom and TA did have a unit cap. TA was like 300 iirc.. and Supcom was 1000.

Huh. Is it per player or the entire game
And what happens when you hit that limit?

Per player and everything stops producing.

I'd point out TA's actual unit cap was 5000 units (the engine limit). This was artificially reduced at launch however later patches removed the limitation.

So, anyone with 290x can just say average fps? pls

For TA:

250 was the original cap, you could raise it to 500 by editing TA.ini.

I believe the mac version was eventually patched (officially) to allow 1000.

There are 3rd party patches that allow 1000, and 5000.

The unofficial TA patch ( v3.9.02) allows 1500, the actual number is configurable in TA.ini.

There is a hard limit of 5000 units in the engine (which is why the 5000 unit patch came in).

5000 units- for a game designed in the mid 90's... it was waaaaaaaaaay ahead of its time

Yep, and even better, unlike some other popular RTSes of the time, a single unit takes up 1 unit slot, not 2,4 or even 8

...is 1500 per player

for 12 players....

....which today it can run stable... and as Cdrkf well said in TA it's 5000 per player. and it can run that stable.

PA has 10 players max and starts crapping out at 5000 units TOTAL. that means if you're in a 10 player game, 500 units per person.

...which is still a glorious sight to behold, mind you.

Just helping putting things into perspective.