multicore or base speed?

[dooom]

which is more important for a10c? should I get a slower clocked quad core or stay with my core 2 duo e8500 at 3.2ghz?

 

more cores or more speed for this sim?

[Griffin]

More speed as in Black Shark. The only feature that takes advantage of multiple cores is the sound engine. So two cores is enough at the moment as there is nothing built into the sim that will use more. For any other modern games, quad core can be useful. I just went from E7300 to Q9550 in the hope of increasing performance with other games. Otherwise I wouldn't had bothered.

Edited December 11, 2010 by Griffin

[dooom]

thanks grif +1

 

that said - what about gpu? i have a 512 9800 gtx... is it time to put a larger card in?

[TeeJay82]

thanks grif +1

 

that said - what about gpu? i have a 512 9800 gtx... is it time to put a larger card in?

 

its time :thumbup:

[Teej]

This game engine is weird.

 

There seems to be minimal benefit to a 4xx card (in geforce speak) over a 260+...AT THIS TIME, and in these sims...

 

GHz does dominate over cores, all else being equal. Between C2D / C2Q, the higher clock speed will probably prevail. However, an i7 at the same clock speed as a C2D/C2Q will outrun it.

 

Based on a small sampling size in terms of # of people I've discussed it with...SLI/xfire seems to be pointless in the DCS engine at this point (at least in terms of frame rate. If you're driving additional displays, that's another matter).

 

Well...not weird...just not optimized for top end technology (cores, sli)

[TwoLate]

This game engine is weird.

 

There seems to be minimal benefit to a 4xx card (in geforce speak) over a 260+...AT THIS TIME, and in these sims...

 

GHz does dominate over cores, all else being equal. Between C2D / C2Q, the higher clock speed will probably prevail. However, an i7 at the same clock speed as a C2D/C2Q will outrun it.

 

Based on a small sampling size in terms of # of people I've discussed it with...SLI/xfire seems to be pointless in the DCS engine at this point (at least in terms of frame rate. If you're driving additional displays, that's another matter).

 

Well...not weird...just not optimized for top end technology (cores, sli)

 

What Teej said is spot on. :thumbup: I had a 260 OVC and a friend gave me a 480gtx and I saw no difference in FPS. 480gtx because of the more video memory did allow me to turn up eye candy without a frame rate loss. Also the comment on CPU is spot on. Good advice there.

 

Memory will play a small part. Faster memory can give a small increase with a cpu dependant sim. Talking about the bandwith of memory.

[EtherealN]

Minor note: multithreading is not technically a question of "optimisation" (though it might be a bit splitting hairs). Basically, optimization is when you take a routine and/or function in the code and streamline it to remove unecessary steps in the code, while multithreading is a much more large-scale enterprise that needs to work at the large-scale architecture of the process.

 

The future will most likely see gradual migration of simulator elements into separate threads as and when they are implemented, sort of like the new sound engine.

 

Also, I want to re-emphasize what others have said regarding clock speed: faster is better, but it only works for comparison between the same architecture - a 3GHz i7 will have considerable advantages over a 3GHz C2D (and it will have free cores for the OS and other things to live in).

[Teej]

Minor note: multithreading is not technically a question of "optimisation" (though it might be a bit splitting hairs). Basically, optimization is when you take a routine and/or function in the code and streamline it to remove unecessary steps in the code, while multithreading is a much more large-scale enterprise that needs to work at the large-scale architecture of the process.

 

As anyone who knows me knows, I'm always game for a bit of semantic hair-splitting, but that's going too far for me.

 

Yes, in one sense (generally relegated to compilers) that's what optimizing means.

 

In a more general sense, and for anyone who lives outside the world of compilers, optimizing means making the best use of resources. Lots of things get "optimized" that have nothing to do with streamlining loops. :D