With 28 lanes, you can run x8/x8/x8 for 3X GPUs and still have 4 lanes open for a PCI-e SSD like an Evo Pro 950. The key differentiator between the i7-5930K and the i7-5820K is number of PCI-e lanes. One other thing, though: Don't bother with an i7-5930K for 3D rendering unless you plan to scale over 3GPUs for your GPU rendering.
#Best printer for macbook pro 2015 Pc#
Render time is a factor of complexity and if you have something that renders 50% faster on a desktop PC than a MacBook, do you really care if your render time is going from 2 minutes to 4? If you're working with smaller, less complex things, maybe it's not such a big deal. It's just not something I would want to do to myself.
For me, moving down to a MacBook Pro wouldn't just double or triple my render time tools would stop working. I like lots of individual light sources, I like lots of highly reflective surfaces, I like SSS. I'm usually rendering in the 2M+ poly range. So, is the MacBook "enough?" Depends on the complexity of what you do and the value of your time. So the hardware's worse, features are less available, and forget about using end points like Octane that require CUDA because you won't be able to run them. Not only is a GTX 970 5x the horsepower of the R9-M370X in a MacBook Pro, but you've got to deal with a) poorer support for Cycles on AMD hardware and b)Īpple's poor implementation of OpenCL in MacOS. On the GPU render front, the MacBook Pro is in really sad shape. Fluid simulation, fabric simulation, collision detection, all CPU-bound, so you're not just getting utility out of those cores while rendering. The other thing to remember is that even outside of render time, all of your simulations are CPU-bound. Speed holding constant and more cores amounts to a dramatic drop in render time. It's not quite a linear improvement but it's truthfully fairly close. The core count advantage will also significantly impact your render times. I've managed an i7-5960X with a stable OC of 1Ghz on 280mm water which you can get in an AIO like Corsair's H110 so it's not out of this world in terms of expense and achievability. You can do the same thing to an i7-5930X. The difference between an i7-5960X at 3.3 Ghz (stock speed across 8 cores) and an i7-5960X at 4.2 Ghz was about a 25% reduction in render time. I did a lot of testing with LuxRender during the development of Reality 4, and since the beta for R4 coincided with the release of Haswell-E, my testing overlapped with my move to an i7-5960X and a GTX 970.Ĭores and clock both dramatically impact render time.
0 kommentar(er)
