Performance problems in the latest version? ATI in particular?

[keith.lamothe]

Ok, both 5.044 and 5.045 contain changes intended to fix the performance issues, please let us know if it helps OR if it does not help

In my case, out of the three saves I tried (DrTall's, Kahuna's, and Varone's save), Kahuna's was the only one where I saw what I would consider "unplayable" conditions, and I'm suspecting that's because it's a fairly intense scenario and I was trying to run it on my craptop rather than a real machine.

Tested the latest changes on Varone's save and I was seeing maybe a 15% improvement, but it wouldn't have been that hard to play even without that improvement, even on my laptop-of-questionable-spec.

Thanks for the help!

[Diazo]

Oops.

Missed this post when I posted in the patch 5.045 thread.

I am not experiencing any of the performance issues that have been reported.

Rig:
i7-920 on Asus P6T SE, 6gig ram.
Radeon 5780 GPU
Windows 7 64bit Home.

All drivers up to date, just checked the ATI drivers less then 2 weeks ago.

D.

[Varone]

Ok, both 5.044 and 5.045 contain changes intended to fix the performance issues, please let us know if it helps OR if it does not help

In my case, out of the three saves I tried (DrTall's, Kahuna's, and Varone's save), Kahuna's was the only one where I saw what I would consider "unplayable" conditions, and I'm suspecting that's because it's a fairly intense scenario and I was trying to run it on my craptop rather than a real machine.

Tested the latest changes on Varone's save and I was seeing maybe a 15% improvement, but it wouldn't have been that hard to play even without that improvement, even on my laptop-of-questionable-spec.

Thanks for the help!

Yeah i notice the improvement too, cheers!

[keith.lamothe]

Yeah i notice the improvement too, cheers!

So... it's ok now?   I don't have to rip out a bunch of code changes to get it to stop doing this?

[TechSY730]

Yeah i notice the improvement too, cheers!

So... it's ok now?   I don't have to rip out a bunch of code changes to get it to stop doing this?

Well, if you want to, you could still rewrite all the innermost loops to be more efficient, twiddle around with data structure layout to improve cache locality, or possibly rewrite performance critical code in hand tuned ASM.
But yea, I don't think it would be necessary.

[keith.lamothe]

Well, if you want to, you could still rewrite all the innermost loops to be more efficient, twiddle around with data structure layout to improve cache locality, or possibly rewrite performance critical code in hand tuned ASM.
But yea, I don't think it would be necessary.

Generally speaking, if performance is ok in version X, but not ok in version X+1, I think in terms of undoing whatever X+1 did to cause that problem, not in terms of brand new optimizations to try to sweep the dust under the rug

[Varone]

So... it's ok now?   I don't have to rip out a bunch of code changes to get it to stop doing this?

I think your off the hook this time Just played an hour or so and everything seemed to run smoothly just like before.