I have a quick question. I currently have a dual Xeon set-up running x5650s @ 2.67. I wanted to know if the latest i7 CPUs that are out now would either outperform my current set-up or should I stick with what I have for now. I currently use 3DS Max 2010 with Rayfire and Afterburn plugins.
I was thinking that having one of the top i7s would have better speed in terms of special effects and high poly processing in the viewport(s) but didn't know if I would take a hit on the rendering side. Any feedback would be greatly appreciated!!
The xeons are still faster, you have 2x6 cores. The best you can hope for in an i7 is 3930k/3960x(i consider them the same cpu) and get about 60% of your current performance
or 3770k<- which is more energy efficient but get about 40-45% of what you have. Its still not worth it.
I am wondering what gpu you have. Rayfire is physX based so there is a chance of improvement there.
If you have a single Quadro card you can offload the physx calculation to a separate card. Doing the compute on the same card that does the rendering can slow the gpu down. (around 10-15%)A medium Geforce like a GTX 460 can do this no problem.
If you have a FirePro(or any non NVIDIA card in the system) then the physX is loaded on the CPU(this is the worst case). I'm not sure how it is in profesional tools(i dont have much experience), but in gaming this meant around 10fps MAX compared to the 40-50 you get doing it on GPU. <- this does not mean it translates the same in a Modeling tool, i believe in these the hit should be much less around 25-30%.
P.S. As i researched about it, i noticed that what i had in mind may differ in scale to your problem, so its best if you give us the full spec of the machine and usage data during load (CPU/GPU/memory)
This is not a game and you cannot have extra physx cards. Rendering is still cpu based with just the extra physics calculations of the plugins offloading to the gpu unless it is baked. In viewports, the card is rendering that as well as physics but max still doesn't support dedicated physx cards.
It might still be useful to know cpu/gpu usage to know how much work is being offloaded to the gpu. It may make you want to upgrade your gpu instead of your cpus.