Hacker News .hnnew | past | comments | ask | show | jobs | submitlogin

Let's say I'm running a post shop and I want to render from After Effects as fast as possible. In normal use, just how much faster will one of these dated Xeons render my comps than the quad-core i7 in a top-of-the-line iMac?


Here is a relevent benchmark. The top of the line iMac has a Sandy Bridge 2600, and the 980X is the desktop version of the equivilant server part.

980X vs 2600 video transcoding: 24s vs 29.1s (lower is better)

But there are still 2 CPUs in the Mac Pro, so you can double performance; it should be 2.5x faster still.

Ivy Bridge, which features in the new MacBook Pros, and likely iMacs soon, nets another 10% or so for the same clock speed over Sandy Bridge.

Haswell will likely be the point at which a quad core Haswell chip roughly equals a 6 core Westmere chip (not bad for a 3 generation gap).

http://www.anandtech.com/show/4083/the-sandy-bridge-review-i...

Edit: Another commenter pointed out that AE has CUDA acceleration. In that case, even a mid range graphics card would be faster than two high end CPUs.


After Effects has NVidia acceleration (CUDA), so a Mac Pro with multiple NVidia cards might be a better choice.

This is an "upgrade" in the sense that the processors they were using became too expensive and they used the currently sold ones.


Depends on how heavily After Effects leverages threads. With up to twelve cores on a Mac Pro, it could potentially be much faster.


Why can't you offload to render nodes?




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: