Comcast’s Wider Pipes and Broadband Caps Seemingly at Odds: P2P Complicates Matters Print

SCOTTSDALE, Ariz., November 4, 2008— Much as we expected, Comcast is responding to the threat of Telco insurgents with much wider broadband pipes. Utilizing DOCSIS 3.0, which allows the cable companies to bond at least 4 RF channels, Comcast is driving toward the goal of providing wideband to 20% of its footprint by year-end.


For More Information or to Buy the Report

 Comcast’s new service comes in two flavors for consumers:


  • up to 50 Mbps of downstream
  • up to 10 Mbps of upstream
  • $139.95 per month


  • up to 22 Mbps of downstream
  • up to 5 Mbps of upstream
  • $62.95 per month


However, one has to start wondering what people are going to do with all this speed. Most online video have bitrates that work with our average 5 Mbps speeds, so 50 Mbps seems like overkill. Assuming that HD downloadable videos from the likes of Xbox Live (6.8 Mbps), Vudu, or CinemaNow (Amazon etc) is the target application, then these speeds could prove useful.

Incongruence of Implementing Bigger Pipes AND Bandwidth Caps

Currently all Comcast residential broadband subscribers are subject to a 250 GB caps. However, if these faster speeds are targeted at heavy users, or those wishing to consume large amounts of content, say HD video, then it would seem logical to increase the bandwidth caps for these users.

There are many hypotheses as to why Comcast would implement wider pipes and bandwidth caps simultaneously. Some assert the bandwidth caps are meant to discourage uptake of video services outside Comcast’s network. Others assert that these actions portend a fully metered service (e.g. pay for bits). With the same caps and faster speeds, the only thing you accomplish is reaching the limit that much faster. Of course, we could point to FiOS and say this is simply a means to remain competitive, in a looming broadband “arms race”. These reasons however are all suppositions that may or may not have merit. One thing is certain, cable providers are trying to deal with the heavy burden of P2P traffic as the networks are increasingly utilized for video.

HD Video and P2P at Odds with Cable Providers

As new video services leverage non-facilities based broadband, the demands on these pipes grows in kind. In addition, the slow migration to higher quality video (e.g. HD) is also driving demand. If we take an extreme case of downloading a high bitrate 1080p movie at 20 to 40 GB a shot, the Comcast 250 Gigabyte broadband data cap does not last long.

Obviously, all HD is not created equal. Many sites and services now offer “HD” quality video, including Hulu, Apple, Vudu, and Xbox Live. We use quotes because in most cases the bitrates employed pale in comparison to Blu-ray or even terrestrial broadcasts. For reference, Blu-ray has a maximum bitrate of 40 Mbps and ATSC HDTV is 19.39 Mbps. Apple iTunes in contrast uses 4Mbps for HD, Hulu 2.5 Mbps for HD, Vudu 4.0-4.5 Mbps for HD and 9.0-9.5 Mbps for HDX, and Xbox Live 6.8 Mbps for HD. Furthermore when we consider P2P, the upstream traffic adds to the demands on these broadband channels. Given asynchronous data transmission, upstream pathways are more prone to congestion, which seems to fit nicely with the larger 5 – 10 Mbps upstream plans.

And yet, enforcing a broadband cap on customers paying $139/month for a blazing fast Internet pipe could be problematic. However, the issue of P2P traffic cannot be ignored. P2P data currently represents 44.0% of all consumer traffic over the Internet and 33.6% in North America. Additionally, legitimate content providers increasingly see P2P as a cost effective way to distribute digital entertainment services and content.

Bandwidth Cap Alternatives

There are alternatives to bandwidth caps. However, they are not without issues either. Comcast, for example, is employing traffic management strategies.

Comcast’s first approach to traffic management drew the ire of the FCC. The strategy used deep packet inspection (DPI) technology from Sandvine to identify P2P traffic and manage its volume. Since the available bandwidth is asynchronous with the upstream being more constrained, this was the sole target for the traffic shaping strategy. In fact, according to Comcast, the upstream data flow was the only source of the network congestion. In addition, Comcast only targeted unilateral upstream data flow. In other words, bidirectional data flow where the subscriber was actively participating in the “transaction” of data was not targeted. When a threshold was reached, the Sandvine unit issued instructions to “reset packets” which delayed unidirectional uploads for the specific P2P protocol in the geographic area managed by the Sandvine unit (Sandvine Policy Traffic Switch 8210).

The new strategy does not target any particular protocol but rather sets two priority levels to target individuals using the most bandwidth. In essence, there are two main states: no congestion, where all subscribers are classified as Priority Best Effort (PBE), and near congestion/congestion, where consumers who exceed 70% of their provisioned bandwidth upstream or downstream will be reclassified as Best Effort (BE). Provisioned bandwidth is determined based on each respective subscriber’s service plans. Data points are taken every 15 minutes and reclassifications are made depending on changes to user behavior.

In effect, this strategy reduces the QoS for those users engaging in high bandwidth activities, such as the downloading HD video. It is uncertain if the 70% threshold will necessarily hold, since cable broadband speeds vary depending on the usage of subscribers sharing the same pipe. During peak hours, a subscriber’s data speed could dip below the necessary 70% threshold – particularly if we consider streaming video at lower bitrates and the 50 Mbps provisioned bandwidth. In addition, video files grow in size (or bitrates increase for streaming), these individuals could find themselves reclassified as BE, when they are in fact engaging in actions well within the intentions of the broadband service. One could argue this is ground to pay for a higher tier of service – or reason to institute a content revenue share with the ISP.

This Intelligence Brief comes from MultiMedia Intelligence P2P Networking: Content’s “Bad Boy” Becomes Tomorrow’s Distribution Channel research. The report addresses the P2P market in terms of worldwide broadband penetration, consumer consumption of data, audio and video files and its associated revenue. Forecasts are provided on an annual or per monthly average (e.g. PB/month) from 2006 through 2012. Forecasts include:

  • Worldwide and Regional Broadband Penetration
  • Worldwide Consumer Internet Data Volume
  • P2P Traffic as Measured by Quantity of Bits Transferred, by Content Type and by Number Monthly P2P Files
  • Total Media P2P Files by Content Type, Monthly Data Volume by Content Type and Legal Media by Content Type
  • “Value” of Purchased and Pirated Content for Music Tracks, Short-Form P2P Video Content, Episodic P2P Video Content and Long-Form P2P Video Content
  • Apportioned US ISP CAPEX and OpEX Resulting from P2P

For more information visit

or contact

Rick Sizemore

(480) 213-4151

This e-mail address is being protected from spambots, you need JavaScript enabled to view it  

About MultiMedia Intelligence

MultiMedia Intelligence, a market research and consultancy firm, specializes on the markets and technologies for delivering and monetizing digital content and services across multiple platforms. We look beyond the classic 'three screens,' which include TVs, mobile handsets, and computers. We put markets into the broader context of the industry ecosystems that are converging and changing traditional business models.