Comments on: WebRTC and Server GPUs? A whitepaper https://bloggeek.me/webrtc-server-gpu-whitepaper/ The leading authority on WebRTC Sat, 02 Jul 2022 17:05:11 +0000 hourly 1 By: Ron W. Szpak https://bloggeek.me/webrtc-server-gpu-whitepaper/#comment-118501 Tue, 03 May 2016 14:27:30 +0000 https://bloggeek.me/?p=10330#comment-118501 Hi Tsahi,

Also lots of potential with low power, C Programmable parallel processors such as Coherent Logix HyperX:

https://www.coherentlogix.com/applications/video-processing-at-the-edge/

http://www.coherentlogix.com/wp-content/uploads/chart2.jpg

Ron

]]> By: Ariel Linker https://bloggeek.me/webrtc-server-gpu-whitepaper/#comment-118500 Fri, 29 Apr 2016 19:00:39 +0000 https://bloggeek.me/?p=10330#comment-118500 In reply to Robert Welbourn.

Hi Robert,

Thank you for your comment.

SURF utilizes the INTEL GPU for its solution at the moment – this might be enhanced in the future to support additional GPU accelerated processors.
Please don’t hesitate to contact me for any additional details you may require on our solution: ariel@surfsolutions.com

Best Regards,
Ariel Linker
Vice President, Sales & Marketing
SURF Communication Solutions, Ltd.

]]> By: Robert Welbourn https://bloggeek.me/webrtc-server-gpu-whitepaper/#comment-118499 Fri, 29 Apr 2016 17:59:28 +0000 https://bloggeek.me/?p=10330#comment-118499 So what is SURF using as a GPU? The integrated Iris GPU on the Xeon E3 series processors? Or are they supporting nVidia’s GPUs? In either case, it would be interesting to see how it affects transcoding performance, for example of H.264/VP8 sessions.

It will also be interesting to see how this works out in a cloud environment. Amazon has support for nVidia-based instances for high performance computing, which would favor a CUDA-based approach, as opposed to using OpenCL.

]]>