Quote: "100 versions of a model is not the same as 100 unique ones of course, instances are always faster. I have to agree with Amen here, at least as a full product goes. Max is ok with a flat empty map and objects, but sculpt and texture the entire terrain and add foliage around and fps drops through the floor"
Shouldn't have to mention for the 100th time, whether or not you are aware that the terrain and grass painting features aren't completed, even the terrain physics aren't working correctly current. So no you can't compare classic and max on incomplete features.
It is in no way representative of performance if you take a incomplete feature and call it proof that performance is worse in max.It's a stupid argument and it has been mentioned time and time and time and time and time and time and time and time and time and time and time and time and time and time again it is a incomplete features. Hopefully that is enough "and times" to let it sink in that you are trying proof performance discrepancies on some thing that is incomplete or doesn't even exist at present.
So no you can't use terrain as proof of poor performance, you can however compare on like for like abilities like indoor environments, pixel shading, mesh handling, memory management, texture management , realtime dynamic lighting and HDR which isnt present in classic but you can however compare the visuals differences between the two.Max has certainly improved on those fronts and much better.
Max is capable of handling more polygons period, more draw calls, GPU memory management is better.Pretty much every aspect of the rendering engine is better and faster period.If it were possible to back port completely populated levels created in max to classic I would bet the performance will be worse. I guess you can also port over levels created in classic to max, however that isn't currently possible, it's likely it will never be 100%.
So no it isn't about disagreement or opinions, it is based on physical and factual data available at present and as such max kicks classic's teeth in, with regards to rendering speed, GPU memory management, mesh management, polygons, texture and multitextures, considering the realtime dynamic lighting system, which classic will not be able to handle period.Basing the argument of poor performance on a incomplete feature as proof is silly, literally makes no sense, additionally being aware that the terrain system in max will be much larger boggles the mind. I suggest we revisit this when the terrain system has been fully implemented and the actual terrain physics has been properly added. Lets not forget the entire engine gui and every thing else is 64bit, larger memory caps for one, with better multi threaded support the overall GPU usage has actually gone down, go figure. I haven't even touched on the fact with 64bit support the 4gig ram cap isn't applicable, which means less reliance on page file requests, and with 64bit support larger chunks of data can be processed it also increases the overall bandwidth between the ram/cpu/gpu. While technically 64bit computing in general would be faster, however being able to more because of it does even things out a bit. which does mean you can erode any gains made as a result. But it does mean overall you can take advantage of that with more complex scenes and faster data crunching overall which may not be necessarily possible under a 32bit application with less resources.
Just the fact that max is 64bit automatically gives it an edge over classic in raw processing power and speed, if managed properly it can indeed be faster but you still end up negating gains made with more complex scenery and data crunching it entails. Which means classic would probably not handle the realtime dynamic lighting and HDR very well, in return the better lighting system and HDR does cost max some performance, the question is whether or not it is worth. The short answer is yes.
Don't believe me, well get GPUZ out and compare two scenes in classic and max.GPUz has the ability to log memory, usage for, system, pagefile, and GPU load, memory controller and if you add CPUz to the mix you can also monitor and log CPU and cache usage as well.
When I hear max terrain performance is worse, when it hasn't even been implemented fully or completed after lee has stated it over and over again it isn't being worked on and several things needs to be completed before the actual backend can be worked on.
This image sums up my general feeling.........when terrain performance is mentioned.
Win10 Pro 64bit----iCore5 4590 @ 3.7GHZ----AMD RX460 2gb----16gig ram