Does RAM make a difference while encoding h.264 ?

New Here ,
Nov 07, 2020 Nov 07, 2020

Copy link to clipboard

Copied

These are my Medie Encoder specs:

Hardware Encoding (Intel Quicksync) enabled.

NVIDIA Geforce GTX 1050  (Mercury playback Hardware Acceleration OpenCL enabled)

I have SSD+1TB hard drive comobo for storate.

 

As an experiement, I tried encoding a 7 minute , 40 second video, at a low Media 720p format.

 

It says that the estimated time will be 2 hours to encode this small video.

 

I have attacked my task manager screenshot for reference.

Can someone help me so as to what I need to do to accelerate the encoding process.  I really want to be able to encode my 10 minute videos in like 1-2 minutes if possible at 4k resolution.  It currently takes me 4 hours to fo 4k videos of 8 minutes. With my current specs.

 

Thanks

 

whitchurch5E6F_0-1604771518218.png

 

Also I see very low Integrated GPU utilization. I was assuming the Encoders get a boost for H.264 content via Intel Quick Sync. More workload shared between CPU and Discrete GPU.

 

Also, why am i getting OPENCL, instead of CUDA suppor for my NVDIA Geforce GTX 1050. That seems strange.

 

Thanks

TOPICS
How to

Views

108

Likes

Translate

Translate

Report

Report
Community guidelines
Be kind and respectful, give credit to the original source of content, and search for duplicates before posting. Learn more
community guidelines

correct answers 1 Correct answer

Contributor , Nov 07, 2020 Nov 07, 2020

Hi Whitchurch,

 

Adobe media encoder is CPU intensive. I have an 8 processor CPU, 500GB SSD and 32GB memory. When encoding H.264 I use 100% of all 8 cores and 16 GB memory. With this rig, it takes what it takes to encode. I hope this helps.

 

Be well fellow puppeteer - platty

Likes

Translate

Translate
Contributor ,
Nov 07, 2020 Nov 07, 2020

Copy link to clipboard

Copied

Hi Whitchurch,

 

Adobe media encoder is CPU intensive. I have an 8 processor CPU, 500GB SSD and 32GB memory. When encoding H.264 I use 100% of all 8 cores and 16 GB memory. With this rig, it takes what it takes to encode. I hope this helps.

 

Be well fellow puppeteer - platty

Likes

Translate

Translate

Report

Report
Community guidelines
Be kind and respectful, give credit to the original source of content, and search for duplicates before posting. Learn more
community guidelines
Contributor ,
Nov 07, 2020 Nov 07, 2020

Copy link to clipboard

Copied

Oh, I should say that the above is for Adobe Premier Pro files which are CPU intensive. Character Animator is GPU intensive and not so much CPU intensive. I need to upgrade my GPU for CH since my encoding runs slow with CH.

Likes

Translate

Translate

Report

Report
Community guidelines
Be kind and respectful, give credit to the original source of content, and search for duplicates before posting. Learn more
community guidelines
New Here ,
Nov 07, 2020 Nov 07, 2020

Copy link to clipboard

Copied

LATEST

Thanks for the feedback.  Especially the part about RAM playing a big role during the Encoding process.  My guess is more RAM means, less need to go to the SDD  or HDD to load stuff into memory.

 

As for the GPU, part, I was reading that NVIDIA GPUS with CUDA support and more VRAM like around 16GB is what Adobe recommended.

 

I did some optimization, to make the machines life a little bit easier.

- I installed the latest NVIDIA drivers and Intel Drivers.  (This gave me the option to use CUDA for my encoding instead of OpenCL for my Nvidia card).

 

- Also after updating the driver for my Integrated GPU on the CPU.  It has given a little bit of a good boost too.

- My CPU has 4 cores,  and what it calls 8 logical cores (4 cores * 2 threads = 8 logical cores is my guess).

 

- My last biggest optimization was moving all my encoding to the SSD, from my HDD.  That reduced encode times from 1 hour to around 21 minutes. So I guess, there is some speedup.  

 

- But yes I agree, keeping the above optmizations and adding in more RAM should help. 

 

Thanks again for your inputs, I apprecate it.

 

Likes

Translate

Translate

Report

Report
Community guidelines
Be kind and respectful, give credit to the original source of content, and search for duplicates before posting. Learn more
community guidelines