I tried to render a project on my Core2Duo 6850 3GHz machine with horrible results. The render time was 24min/frame where as the same file on my old AMD 4400+ took only 5min/frame. Is AM15 optimized for Core2Duo? What happened to the multi-core setting? Is there some secret to why my DDR2 3GHz machine renders 5 times slower than my old 4400+? What am I missing? Does this program even know what a Core2Duo is? Please help as I need to get this project done ASAP.

Core2Duo 6850 3GHz
X38 Chipset
DDR2 1066
10K Raptor

Multi-core rendering doesn't work well on all projects.

Youre most reliable bet is to run two instances of A:M and render half the frames on each simultaneously

you could set one instance to start on frame 1 and step set to 2 and set the other instance to start on frame 2 with step set to 2. Odd frames and even frames.

This scales well and will get near halving of total render time.

(presuming you have enough memory to support 2 A:Ms).

That makes sense but could someone explain why the much faster Core2Duo renders literally 5 times slower than my old 4400+? Is there some form of optimization that I'm missing? Is AM not compatible with Core2Duo? It's very annoying to think that my newer faster machine is 5 times slower at rendering than my old one. It's not the same on other software packages.

I'm not experiencing any issues on my Core 2 Duo. How much RAM are you running (old machine and new)? Why would you expect AM to be optimised for just one processor type in any case? 50% of users here use Mac's - but I'm no programmer so maybe you can optimise separately.

Playing devils advocate here, if your project is that urgnt then perhaps you can bring your old faster machine into the rendering pipeline and lodge a report on AMReports with all the relevant details. Complaining about it here isn't going to get any potential problems fixed (or your project rendered any quicker).

Cheers

Let us know if the information in this thread helps you:

http://www.hash.com/forums/index.php?showt...mp;hl=registery

Also, I assume you have your OS adjusted to take account of the second core. Apparently Windows has 1 core as the default. Here's how you can change it in vista:

http://www.youtube.com/watch?v=GOVN2KB78bI...feature=related

If not, then please report to AM:Reports. Having the renderer take advantage of newer hardware (even discounting the extra cores) is in everyones interest. Including Hashs'.

Here's discussion on why multi-threading is disappointing on many apps, not just A:M.

http://www.hash.com/forums/index.php?s=&am...st&p=254379

Complaining about it...no. Inquiring about it...yes. Both machines have 4GB ram and I've even cut back to 2 just to check if its a ram problem. The fact that the Core2 is 5 times slower than the other worries me is all. I can't seem to figure out the problem other than I have some setting or other wrong. It's an identical file and a unique load on each machine. No other apps exhibit this sort of problem. It really seems to bog down while rendering the patches.

You've obviously checked that your render settings were exactly the same (like same # of passes?)

Also what about your cache size differences? and virtual memory allocation? and OS's are the same ? or is the core2duo Vista? - the other XP?

Yes, the Vista & XP-Thing can be a huge factor... vista is much slower than XP is...

One other thing:
AMD-Processors were often faster with A:M than intel once were. I dont know how true that is till today, but it used to be that way because AMD has another way to built their processors. So even if the intel-one is faster with other programms, it isnt faster with A:M.

But A:M werent at any time 5x faster than intels...
So I guess that you have a setting which will slow down A:M. For example you use Thread: Auto in your registery (which doesnt work to well for me neither) or you have some other circumstances like Vista installed, etc.

*Fuchur*