v0.31 Talk:System requirements

I couldn't find the Mac OSX version of DF2010. Is there one? --Calculator 17:29, 7 April 2010 (UTC)

Not yet. After a few bugfix versions, (and obtaining some new RAM for his computer), Toady One will of course release versions for other OSes, but for the initial wave of bugfixes, he wants to limit it to one OS, for simplicity. I'd say between two weeks and a month before we see OSX and Linux versions. --Zombiejustice 18:09, 7 April 2010 (UTC)

I am trying to run the new version, but the game is painfully slow (especially compared to 40d). My computer isn't that out of date, so I was wondering what specs are recomended? --Frewfrux 22:43, 7 April 2010 (UTC)

Can't really give any solid recommendations, but I can share my findings between my machines. A 3.0Ghz P4 w/ 1GB of RAM ran the new version a little slower than the old. Another machine with a dualcore 2.0Ghz processor and 2GB of RAM ran it even slower. And my main machine, a quadcore @2.3Ghz with 3.25 GB of RAM ran significantly faster than the old version. These very simplistic tests were done with the same world --Kuroneko 23:45, 7 April 2010 (UTC)

Hmmmmm. Maybe it's the RAM that makes the difference. And then, perhaps it's that I just keep not thinking of my computer as old. The processor is okay, AMD 64 2.2 Ghz, but I only have 1 GB RAM. It ran the 40d version great, except for the very, very full fortresses that were by a canyon. I wonder if anyone else can share their specs, and then maybe we can know what affects performance the most? Just a thought. --Frewfrux 03:19, 8 April 2010 (UTC)

I've been curious to see what affect CPU cache sizes have on the game. It is painful but playable (I get bored at work) on my work PC which is the listed 1.4ghz tualatin (mind you, the game is only getting about 75% of the cpu cycles), but on my home PC (Q9550 @ 3.5ghz) is absolutely screams. If I get the time, I may try running the game on my old P4 box using multiple cpus at the same frequency but with varying cache sizes to see how much of an impact that has. I know the game isn't SMP aware (correct?) so a single fast core will outgun two slower cores in the game. As for RAM, I've seen the game use upwards of 800MB while world generating and under normal game play, for me, it uses between 400-500MB. So you may want to make sure you aren't dipping too heavily into your swap space when playing. That will definitely slow things down. --Helf 03:11, 14 April 2010 (UTC)

It might be helpful to add the fps achieved with each system. Although these are influenced by several settings, fortress size and location they might give a more objective rating than terms like 'slow', 'slowish' and 'somewhat slow'. --Hermano 12:29, 13 April 2010 (UTC)

The push for specifics

Please add more detailed information on the population and landscape of your fortress, like we have on the 40d page. It's nice that you're getting 100 FPS, but it helps to know how much is on your map when that happens. 128.151.186.179 04:33, 18 April 2010 (UTC)

Seriously?

The "comparisons" on this page are absolutely worthless in its own right. No details at all as to the forts in question and such. Seriously folks, you can do much better... - Vrga - Splaterrer wannabeeeee 04:32, 1 July 2010 (UTC)

Then please help improving it. =) --Used 08:37, 1 July 2010 (UTC)

Will try :D -Vrga 10:13, 1 July 2010 (UTC)

Okay, so my idea for the report to use when posting info here.

Game info

Game version:

World size:

Embark size:

Number of dwarves:

Average fps:

Default/nondefault raws:

Tileset in use:

Amount of stone dug:

Amount of water and state:

Approximate amount of z-levels:

RAM usage of game:

Draw mode in init.txt:

PC info

CPU:

MBO:

RAM:

GPU:

OS:

Here's mine for example.

Game info

Game version: 31.08 - graphics

World size: Medium region (default, nonmodified)

Embark size: 5x5

Number of dwarves: 140

Average fps: 6-14

Default/nondefault raws: nondefault, dwarven speed modified, some smelting recipes modified.

Tileset in use: Phoebus prebuilt version

Amount of stone dug: unknown

Amount of water and state: A brook, dammed and diverted into big cisterns for trap use, mostly static and pressurised.

Approximate amount of z-levels: Beyond 200 below the ground level, around 5 upwards from ground.

RAM usage of game: Around 1.25 GB when playing

Draw mode in init.txt: VBO

PC info

CPU: AMD 7750x2 BE @ 2.7GHz

MBO: Gigabyte GA-MA78G-DS3h

RAM: 4GB DDR2 800

GPU: XFX GTX260

OS: Windows 7 x64 Ultimate

Tell me if i'm messing something up... - Vrga 10:25, 1 July 2010 (UTC)

Since no response has been given, i'll be putting this on the main article for people to fill out. - Vrga 22:01, 2 July 2010 (UTC)

Aaaand done... I think i did an okay job. Any reccomendations are more than welcome. - Vrga 22:43, 2 July 2010 (UTC)

Perhaps it would work to make a standardised benchmark save, with a middling population, moderate water features and average elevation changes. This could allow different systems to be tested with the same map and fort,and allow each submission to forego the population and water info each time. Going a step further, specify that it is to be shipped with a vanilla DF, to remove the graphics variables. I'm not sure how compatibility works when moving saves though, and game updates may invalidate the recordeddata to some extent.58.170.59.138 11:11, 5 November 2010 (UTC)