tag:blogger.com,1999:blog-12074752.post4244979365015108087..comments2009-08-20T16:59:36.951+02:00pvehttp://www.blogger.com/profile/08966586301980201453noreply@blogger.comBlogger2125tag:blogger.com,1999:blog-12074752.post-67472455044553344522009-08-20T16:59:36.951+02:002009-08-20T16:59:36.951+02:00

15 hours every minute implies 9000 concurrent streams (assuming upload speed equals download speed). Downloads are 10 times that, so 90.000 concurrent streams (and now we no longer have to assume that upload speed equals download speed). <br /><br />9 videos a day at 5 minutes each equals 45 minutes of viewing time per user. On a user population of 3.2 million you would then have 100.000

pvehttp://pve.myopenid.com/noreply@blogger.comtag:blogger.com,1999:blog-12074752.post-38462910179813689712009-07-28T19:49:42.443+02:002009-07-28T19:49:42.443+02:00

90000, are you on crack? Don&#39;t know where your calculation is based on, but it is hogwash. I measure the average traffic caused by YouTube on our core to customer routers (for a big ISP in The NL). This amounts to averages of each internet user to see 9 YT videos on a slow day (based on bandwidth:videolength:videofilesize ratio in timeline of our measured traffic). Extrapolating those using

Firehorsepowerhttp://www.blogger.com/profile/13561048675649796505noreply@blogger.com