I'm interested in using one of the following for transcoding hi res material into low res proxies:
Intel® Atom™ x7-E3950 (4 x 1.6 / 2.0 GHz, 2MB L2 cache, 12W)
Intel® Atom™ x5-E3940 (4 x 1.6 / 1.8 GHz, L2 cache 2MB, 9W)
Intel® Atom™ x5-E3930 (2 x 1.3 / 1.8 GHz, L2 cache 1MB, 6.5W)
We'd like to encode as H.264 or H.265 relying on the Quick Sync technology that's part of the 9th Generation Graphics.
Other than the clock rate & the number of cores, I'm trying to find out what performance differences there are between the Quick Sync blocks in these processors. Are there any real world benchmarks?
Again, focusing on the Quick Sync block, I'm also curious to know how many video encoding streams each atom could handle concurrently.
I'm also curious to know what the encoding performance difference between a new generation Xeon Core with embedded FGA (I understand Intel Media Server Studio supports transcoding on the FPGA via OpenCL) & a Quick Sync enabled Atom (I know the comparison sound ridiculous) would be.
Finally, I understand that at one point Intel produced the E600C family of Atoms with embedded FPGAs. Is there a current equivalent? Would that be faster than Quick Sync?
The Intel Media SDK provide sample applications that can be used for transcoding to provide real data application.
I am currently investigating if there is any Intel official benchmarking information for those specific processors that you are inquiring about.
Regarding Xeon processors you can check on the following page: https://software.intel.com/en-us/intel-media-server-studio https://software.intel.com/en-us/intel-media-server-studio , this image is from the mentioned page:
Sadly, I don't currently have access to any hardware with which to run benchmark tests for myself.
Thanks for pointing me at the Intel Media Studio Server page. I'd come across that chart before, but from what I could see, it's not particularly clear whether that benchmark is using Quick Sync and/or the FPGA to transcode with.
Thanks again for your quick response.