Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[monolar]

Thought this to be funny

http://episteme.arstechnica.com/eve/forums/a/tpc/f/39309975/m/500002266931

DF as a benchmark tool is actually a cool idea

[Mephisto]

People don't give a hoot about Crysis anymore. It's all about "Can my PC run Dwarf Fortress at more than 10FPS?"

I'm not sure if that's a good thing or a bad thing.

[Fieari]

Someone should upload a save file containing a high population catsplosion (not the thermonuclear kind) and use that for the benchmark.  Bonus points if there are lots of waterfalls over multiple Z-levels.

[Warlord255]

Someone should upload a save file containing a high population catsplosion (not the thermonuclear kind) and use that for the benchmark.  Bonus points if there are lots of waterfalls over multiple Z-levels.

I'd think that a benchmark for performance would be FPS of 7 Dwarves in a 5x5 flat embark zone.

[woose1]

Someone should upload a save file containing a high population catsplosion (not the thermonuclear kind) and use that for the benchmark.  Bonus points if there are lots of waterfalls over multiple Z-levels.

I'd think that a benchmark for performance would be FPS of 7 Dwarves in a 5x5 flat embark zone.

Pff... I can get that blindfolded!

wait...

[LumenPlacidum]

Someone should upload a save file containing a high population catsplosion (not the thermonuclear kind) and use that for the benchmark.  Bonus points if there are lots of waterfalls over multiple Z-levels.

I'd think that a benchmark for performance would be FPS of 7 Dwarves in a 5x5 flat embark zone.

According to my screenshot (partial print on 5 frames before not redrawing), I can manage ~1250 FPS under those conditions.  There were 18 total creatures on the map and no features other than flatness.

Without partial print, I can get ~1000 FPS.

[azrael4h]

Mine usually starts slowing down at around 60-80 dwarves, and 20 pets or so if I have a lot of orders being carried out. I'm thinking of trying to use several mason shops doing nothing but building blocks, with stone blocks disallowed from my ore/coal stockpiles to cut down on hauling jobs. It'd be some really messed up workshops though, with hundreds of blocks.

[Tormy]

Thought this to be funny

http://episteme.arstechnica.com/eve/forums/a/tpc/f/39309975/m/500002266931

DF as a benchmark tool is actually a cool idea

Actually, the OP's post makes sense.
Bombcar QFT:
"Dwarf Fortress wants the biggest, baddest processor you can throw at it. Nothing else matters, and it is single core. A netbook is probably not where you want to look."

[Freaky]

I was just thinking of something like this; I'm looking to upgrade my old Opteron 185, and considering either a Phenom II or Core i7.  Obviously, I'd like to know if the i7 would be worth it for DF, not just silly little games like Crysis and SupCom.

So, we need a relatively complex save with plenty of dwarves, animals and flows, and some reproducable means of getting an fairly accurate FPS reading out of it.  And, of course, someone with an i7 920 and a Phenom II 940

[MuonDecay]

Someone should upload a save file containing a high population catsplosion (not the thermonuclear kind) and use that for the benchmark.  Bonus points if there are lots of waterfalls over multiple Z-levels.

I'd think that a benchmark for performance would be FPS of 7 Dwarves in a 5x5 flat embark zone.

Benchmarks are typically measurements of performance under precise and demanding conditions.

So the only properly empirical benchmark would be a fully self-contained install with preset settings, containing a savegame at a resource-intensive location.

[Freaky]

A friend's just got a 2.66GHz Core i7; I, of course, zipped up a somewhat slow fort and had him see what FPS he managed.  Not a comprehensive benchmark, but simple to do; unpack, run dwarfort.exe, hit enter twice, hit space, note the FPS in the corner.  The results:

• 2.6GHz Opteron 185: 25-28FPS.
• 2.6GHz E6700 Core 2 Duo: 40-45FPS.
• 2.6GHz Core i7: 55-59FPS.

So, the C2D's about 1.6x faster and i7's about 2.1x faster than a first generation dual core Athlon 64 at the same clockrate.

[Kardos]

You also have to keep in mind that an older computer will have a lot more background noise in terms of processing, hard drive access (read/write), memory and general cleanliness (dust on the hardware).  Three brand new computers using identical parts except for each of those processors might (and probably will) show very different results to those you posted.

[Ignoro]

Already suggested this myself in the suggestion forum.

[MuonDecay]

You also have to keep in mind that an older computer will have a lot more background noise in terms of processing, hard drive access (read/write), memory and general cleanliness (dust on the hardware).  Three brand new computers using identical parts except for each of those processors might (and probably will) show very different results to those you posted.

The problem there is that those three different processors are not all the same socket, so that's impossible.

Different motherboards would also be required.

[Shurikane]

Question.

How well could DF run while under the famous nitrogen-cooled CPU that busted something like 7 GHz?