You will not have a noticeable benefit. When i changed the 1066 MHz with 1333 MHz memory, Geekbench showed maybe 1% more performance, or more probable it was only minimal data fluctuation.
For example, the difference between DDR3 2133 MHz and DDR3-1333 MHz is less than 2%, and there is 4.5% difference from DDR3 2133 MHz to DDR3-1066 MHz.
I quote:
http://www.tomshardware.co.uk/forum/314892-30-cl11-what-difference
http://www.computerbase.de/2012-05/test-welchen-ram-fuer-intel-ivy-bridge/3/
Sorta figured as much, but was looking to verify the hunch. Thanks.
I'm kinda shooting for the low hanging fruit. First the GPU's as this box is a render / Rhino machine. Pair of GTX970's should address that.
Drive speed next, Samsung 1TB 850 EVO addresses that. I'm sort of curious as to why a lot of you folks are running the SSD's on PCIE cards rather than on a bracket in a bay. Is there that much of a performance difference?
I'm not looking to do video editing or anything that needs raid level performance for large storage (although I do have an ESata external Raid 5 for my photo work), just modeling and rendering frames, so I figure the internal bay route is fine, as the two GTX's will eat 4 slots, and I have the ESata card for the raid in the 5th.
Next the procs, a pair of Xeon 5690's should address that.
Box has 12GB, need to update that, but wanted to ensure I went the right route on that end which your answer addressed nicely. Thankee.
It's been a bunch of years since I've had to milk the sweet spot of price / performance out of a Mac box. I had kinda forgot how much 3D work is just as much about wringing performance out of the hardware as it is the business of checking continuity of nurbs curves.
BTW, I'm pretty impressed with the Mac Beta of Rhino.
It's still got a ways to go (help system documentation, some tools still not ported, etc) before it's ready to ship, but for the most part is pretty solid.
I'm using Rhino for the modeling, and Blender for the render engine, mostly because I'm used to cooking materials with shader nodes and it's less of a learning curve and cost than a 3rd party renderer.
The work is modeling, short anims & stills for proof of concept / approval checks, then eventually local 3D Prints for STL export checks prior to sending off to a service bureau for final prints for casting. I knew I had to do something when the first 10 second anim took 3 days to render.