I saw some comments about using a software called Large Address Aware in the roll call forum.  I thought it appropriate to take this to a new thread.

Some of the posts said the software was necessary to use 8 cores.

I have 24 cores.  Two E5 cpu's that are 6 core each with intel hyperthreading.  Poser, Vue, autodesk, Terragen 2 and above, DS, all see and use the 24 cores.

TGclassic and Bryce do not.

Bryce will see 8 cores bot only if you set the render priority to high. If you leave it at the medium default it will see half your cores up to 8, so if you have 16, it will see 8. If you have 8 it will see 4.

Now that said, cores are not all the same.

Bryce sees actual cores at about 80 percent and virtual cores at about 10 percent.

Bryce also sees only one CPU.

So for me, Bryce uses the first 6 actual cores at 80 percent and two virtual at 10 percent and the rest of my resources are on vacation.

I am curious as to why "Large Address Aware" is any different than just changing render priorty to max?

By the way, AMD is different.  If you have an AMD 8 actual core cpu, (and yes, I know amd doesn't have hyperthreading), then you will render faster than a 4 core intel with hyperthreading to 8 cores.

I have a formula that shifts a bit depending on the software, but for Bryce, an AMD 8 core at 4ghz would be an index number of 32 (8x4). 

Might as well ignore the virtual cores of intel.  8 core intel at 4ghz would be the same but if you look, the cost of an 8 core (16 total) at 4ghz, is about 10 times the cost of the AMD chip.

Methinks a high speed amd machine with bryce will be cheaper and out perform intel on a price comparison basis.

But......

......

How does the "Large Address Aware" software change this formula?  Will it let Bryce see both CPU's and use them?  Will it let Bryce use virtual cores to full capacity?

I'd love to hear more.