I’m currently looking into / researching getting a second machine purely for rendering and got three ideas in mind but need some thoughts on which is best in terms of rendering speed:
HP Z600 workstation - Dual Xeon X5660 /16GB RAM
Workstation specialist - RS-D2850 - Dual Xeon E5 - 2640 / 16GB RAM
Home build I7 3930k (overclocked)/ 16GB RAM
I’ll mostly be rendering with Maya and V Ray
Any thoughts would help clear my mind ?
Thanks
price might be a big factor as will how your renders are set up with threading.
the two xeon systems will perform similarly. They both have 12 cores. One has newer xeons, but slower clock speeds.
The dual Xeon E5-2640 system will easily be the fastest. The Z600 is using older Xeon processors that are nearly three years old and wouldn’t compete with current generation processors very well. Even if overclocked, with the 3930K you’d probably be getting just a little more than two thirds of the performance of the Xeon E5 system and that would be lacking the ECC memory which you might want if you’re going to be rendering something for days. On the other hand it would be way cheaper too.
you could probably buy 3 or 4 of the 3930k systems for the same money
This website is somewhat interesting for comparing CPU’s. At the very least they lay out all the features:
http://cpuboss.com/cpus/Intel-Xeon-X5660-vs-Intel-Xeon-E5-2640
Hi guys,
Appreciate the advice.
I’m guessing the dual E5-2640 would be around £3000+ ( a bit out of my budget).
I’ve seen an HP Z600 for £1800
Ideally I was hoping for just a single machine as I don’t have the room for 3 or 4 towers
Looking at that website and CPUbenchmark.net the dual CPU Z600 is only marginally better than the 3930k.
Which kind of confused me a little as I assumed using a dual system with 12 cores would be a lot quicker than a single CPU with 6 cores at rendering (even though the 3930k is newer tech).
it’ll depend on the test/benchmark being used. Those particular scores don’t tell the whole story for every scenario, just general use for common computer tasks.
In a perfectly multithreaded world, the dual xeons in either case would be quite a lot faster. Not everything utilizes all cores perfectly though.
However for rendering, in general the dual xeons would be faster than the 3930k by a considerable margin - around 30% faster for the x5660’s
though the 3930k would be about 30% faster at single-core functions which is what are used the majority of the time
If you have room for a full tower dual socket workstation you have room for a few single socket render machines. Some example hardware is below.
http://www.newegg.com/Product/Product.aspx?Item=N82E16813157283
http://www.newegg.com/Product/Product.aspx?Item=N82E16811112300
Assuming the software licenses are cheap or included the best bang for the buck will be with more machines using lower end hardware versus a single machine with higher end hardware.
This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.