4

I'm planning to get a new (2nd-hand/refurbished) system for rendering large POV-Ray scenes. I'm using very large bitmaps as bump map and colormap. I'll usually be running multiple POV-Ray processes at a time to utilize more processor power. (A large portion of the render time is used by single threaded parsing.) At peak moments the memory usage will be above 32GB so I'm aiming at 48GB of memory.

I could choose between a HP Z600 or Z800 with both have almost the same specs (dual Xeon 5570, 48GB) except for the memory speed. The Z800 has 1333MHz memory and is €200 more expensive. How much difference would it make for my workload?

2

A general rule of thumb is that a 100% increase in memory speed corresponds to a 5% increase in overall performance. This is because memory access is slow relative to the CPU, so CPUs and programs are designed to minimize memory access.

My experience with POV-Ray bears this out: overclocking the RAM/FSB of my computer from 266 MHz to 333 MHz resulted in benchmarks taking about 1% less time -- though as you can probably tell from the memory speeds involved, this was ten years ago.

| improve this answer | |

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.