EC2 for video coding

I have a potential job that will require me to do some video encoding with FFMPEG and x264. I will have a series of files that I will need to encode once, then I can knock down the instances. Since I'm not sure about using x264 and FFMPEG resources, what instances should I get? I think that

Extra large instance with lots of resources

7 GB of memory
20 EC2 computing units (8 virtual cores with 2.5 EC2 computing units)
1690 GB instance storage
64-bit platform
I / O Performance: High Level API Name: c1.xlarge

or alternatively

GPU Cluster Quadruple Extra Large Instance

22 GB of memory
33.5 Computing Units EC2 (2 x Intel Xeon X5570, Quad-Core Nehalem Architecture)
2 x NVIDIA Tesla "Fermi" M2050 GPUs
1690 GB instance storage
64-bit platform
I / O performance: very high (10 Gb / s)
API Name: cg1.4xlarge

What should i use? Does x264 / FFMPEG improve with faster / more processors, or does it really increase the GPU pound? In any case, it seems that the Cluster GPU seems to be a higher performance instance. What should I prefer?

+4
source share
2 answers

Ffmpeg recently added support for VAAPI and VDPAU, but this only allows the GPU to decode H.264 video. It uses a CPU for encoding.

+3
source

You are probably better off using zencoder.com, they have a great API, and the quality you choose is likely to be better than hours of work with optimizing Ffmpeg settings.

-3
source

Source: https://habr.com/ru/post/1384415/


All Articles