I don't know if anyone else has measured this, I've only gotten hard #s on two different evenings. . .but. . . I have a Premiere, Vizio smart TV, and Apple TV all plugged a connection with the exact same bandwidth (inside my own house on just this LAN segment in question, I can transfer files from my Mac to my NAS at 128Mb/s, so controlling for other things both my router and my downstream internet are very capable). Comcast is my ISP (& DNS), and though some shows/movies on Netflix are labeled on screen "SuperHD" yet when I go to the test URL in my browser, I get the "not supported yet." screenshots: Yet of all 3 devices plugged into the same connection, there is a pretty significant difference in the bandwidth they can pull down from Netflix. Monitored by the traffic flow meter in my Asus RT-N66U router which averages over trailing 3-5 minutes, and there was no other activity on my network pulling from the internet at that time (I can verify in my lists of live connections, etc.). 1st test on 9/12/13 late evening (tested House of Cards, as screenshot above "Super HD" though info conflicted on that, minimum I should be at least HD (according to what Netflix says I'll need). - 460 KB/s average over trailing 5 min, which is about 3.5 Mb/s to a Premiere in its Netflix application - 950KB/s which is 7.125Mb/s, on the ATV-3 so slightly more than twice the bandwidth on the AppleTV3 (tests were within 30 min of each other as well) two footnotes 1. My Netflix preferences on their website are for highest quality available 2. My Premiere to Premiere txfrs on my network are pretty fast, from 40Mb/s on the slowest one, to 60-80Mb/s if one or both are wired, so my actual Premiere networking appears to be in decent health. 2nd test 9/13/13 late evening, same show, house of Cards - .92 Mb/s streaming to TiVo Premiere (!!) - 2.67 Mb/s Netflix streaming direct to Vizio smart TV - 5.37 Mb/s Netflix streaming to ATV3 Yes differences in video quality were noticeable, but at the same time I was also impressed with the relatively good picture at 1Mb/s. However, this is not good! Here is a sad screenshot in which you can see the difference when I switched from ATV3 streaming to the Premiere. The tan bump of upstream activity was accidentally turning on a torrent server on my NAS for about one minute, even if you canceled out the slight downstream it brought with for that minute, the ATV3 to Premiere streaming switchover is rather conspicuous. The only thing I haven't done in my home is test different Premieres. . .not that I think there will be a significant difference. I just wonder who to blame, and how/why it does this, and if there is anything we can do about it. I'm wondering if anyone else has tested (controlling for other things like speed/connection quality, other LAN traffic, on multiple nights, same show, etc.).