Yeah I think I support that recommendation. You might want to check the price difference between the 4770k and 4790k they are almost identical performers. Maybe consider also the differences H87, X87, and H97 ATX chipsets. If you are not significantly overclocking, running multiple GPUs, or add-in boards that exceed the H87 PCIe config, then the z97 is over priced. Chances are you will be using the integrated graphics over your HD4000 series GPU you have now. The HD4000 will pale in comparison to the HD4600 with quicksync, opencl1.2, and AVX2 support. With no discrete GPU and standard sata storage, the z97 is overkill. The H87 will allow you upgrade to a single PCIe 3.0 x16 GPU in the future. By the time you upgrade again, you will most likely want DDR4 architecture.

Xeon is generally looked to for more threads(CPU cores,) greater memory size, and bandwidth in the way of more memory channels. The x99 chipset looks more like a Xeon class machine with not only DDR4 increased bandwidth, but 4 memory channels, and more actual CPU threads.

I forgot to mention my graphics card is a GTX670 but not sure that matters. The Xeon apparently is an i7 4770 minus the graphics and the OC ability. I'm not set on the Z97 Pro but it's not terribly expensive, not that I wouldn't mind a lower price board either.

I wasn't sure if an OC'd i5 would trump the Xeon X3 1231 in rendering times and that's where I'm not well versed in. Does it come down to pure speed or hyper threading, or both?

I see what you mean about the Z97 board, unnecessary expense if I cannot OC or use SLI. I may change the MB preference to the H97 plus...makes sense.

Option C is to go for only a CPU upgrade to the AMD FX-8370. As far as rendering goes, has this been a proven CPU or have users reached for the i7 instead?

If you are concerned with the number of threads available (and Vegas can use 16 readily), then you want to go to Intel. The AMD's are 1 thread per core, the Intel i7s are 2 threads per core. The i5 is not hyperthreaded, that is, 1 thread per core.

@dka56

Here is why I do not prefer AMD FX series CPUs. I think the APU series is a step in the right direction, if AMD simply increased the size of the APU such that they could offer an R9 or better, they would be in business. The A10 with the R7 GPU is close, but even that system is basically an i7-800 series with a sub hd5700 series. Basically it compares to a state of the art system in 2009, which would be fine if all you want to do is some light HD video editing of AVC content.

The second reason is this:

Chipset

Hypertransport connecting the southbridge is barely enough bandwidth for all the devices they are putting out there. This may also be the reason they are still on PCIe 2.0, since 2.0 has 20% overhead and roughly only 6GB/s throughput. One PCIe 3.0 x16 slot would leave about 5GB/s for all sata, audio, other PCIe slots and overhead. 2x PCIex16 slots would serious over subscribe the HT bus without additional buses, which we do not see on the diagram. I just do not think you can move enough data between memory and display compared to the way Intel is optimized.

Even DMI 2.0 on the Intel side is in jeopardy of being over run with SSDs in a raid configuration. Adding a separate x8 controller on the right chipset, would be the only way side step this bottleneck on desktop motherboards.

Based on research I have done, AMD GPUs have better support for OpenCL as a standard, and Vegas Pro is trying to leverage OpenCL for acceleration. When optimizing a system for Vegas, look for the best computational scores in Single precision for both CPU and GPU combined, then you need be able to throw enough pixels to screen for the size of your display, then select the widest/fastest transports between system memory and display screen. That means multiple memory channels at high frequency, high CPU frequency which equals GB/s, PCIe3.0 x16 GPU interface, Display port 1.2 or better interface on monitor and GPU.

Here is an interesting post on 8-bit mode vs 32bit FP which I believe supports single precision score of GPU and CPU theory.

Currently I think the best of both worlds when optimizing a system for Vegas Pro use, would be a 16 core Xeon with an AMD W9100. For the rest of us an I7-5xxx or i7-4790 with an R9 290x and 32GB of fast memory.

If we all do not like the stability and issues with GPU acceleration, we could go back to custom video hardware that changed often and cost thousands. I think what OpenCL offers is amazing when compared to dedicated video codec cards of the past., that only handled one codec.

Astar, great post....really puts things into perspective......thanks.

I think the only reason I would go with a Z97 is if another CPU became available down the road and needed the flexibility. However the Asus H97 plus looks like a good deal at $100.

As far as memory, going with the Xeon X3 1231, would my 4x2GB Corsair Dominator 1600 be less than optimal? I don't know how memory will change rendering speeds to the point I will notice larger, more pronounced lag, but if I'm going with a new MB, CPU, then the RAM is on the chopping block if it needs to be.

While rendering a HD video now on my current system, I never see the RAM usage really break the 4 GB barrier.

Edited to add, I updated my system specs.

From Old Smokes earlier post......

How would the i7 4790 compare to the Xeon? Granted this is not the unlocked version. I agree the 4790K would be a better choice but that is taking the price a little high. I would like to stay around $250 but can move a little.