- Mar 4, 2019
- 1
- 0
- 10
HI,
We are working on a OTT/IPTV streaming video solution. I feed IP streams from a hardware encoder in H.264 HD and SD MPEG 4 format, mostly channels are in .MP4 container format. But to transmit on internet/Intranet and at different bit rates, i need to transcode in case of SD MPEG 4 to H.264 Codec and in case of HD H.264, I need to transrate the streams, I have to manage 100 such streams( 70 SD and 30 HD), I i use CPU, it is getting overloaded with 12 channels itself.
Hence, I am looking for graphics card, as per Wowza - Nvidia the below link supports .https://developer.nvidia.com/video-encode-decode-gpu-support-matrix https://developer.nvidia.com/nvidia-video-codec-sdk The first is a matrix of all of their GPUs which lists the capabilities of each one. The lists are broken down into Encode (NVENC) capabilities and Decode (NVDEC) capabilities. I need to use "Unrestricted sessions" GPU as multiple streams are required. Depending on Maxwell, Kepler architecture, the number of streams that can be played is decided. I am not sure should i go with Quadro OR Tesla.
If there is any link that gives Number of streams that GPU can encode/transcode/decode, then it is easier to decide based on cost involved.
I need to place this cards in a server - Also, can a combo of Quadro and Tesla be kept in 1 single server
Any immediate response is highly appreciated
Regards
Suraj
We are working on a OTT/IPTV streaming video solution. I feed IP streams from a hardware encoder in H.264 HD and SD MPEG 4 format, mostly channels are in .MP4 container format. But to transmit on internet/Intranet and at different bit rates, i need to transcode in case of SD MPEG 4 to H.264 Codec and in case of HD H.264, I need to transrate the streams, I have to manage 100 such streams( 70 SD and 30 HD), I i use CPU, it is getting overloaded with 12 channels itself.
Hence, I am looking for graphics card, as per Wowza - Nvidia the below link supports .https://developer.nvidia.com/video-encode-decode-gpu-support-matrix https://developer.nvidia.com/nvidia-video-codec-sdk The first is a matrix of all of their GPUs which lists the capabilities of each one. The lists are broken down into Encode (NVENC) capabilities and Decode (NVDEC) capabilities. I need to use "Unrestricted sessions" GPU as multiple streams are required. Depending on Maxwell, Kepler architecture, the number of streams that can be played is decided. I am not sure should i go with Quadro OR Tesla.
If there is any link that gives Number of streams that GPU can encode/transcode/decode, then it is easier to decide based on cost involved.
I need to place this cards in a server - Also, can a combo of Quadro and Tesla be kept in 1 single server
Any immediate response is highly appreciated
Regards
Suraj
Facebook
Twitter
Instagram