PeerApp, Pando Collaborate to make Peer-to-Peer Palatable to ISPs
When it comes to peer-to-peer (P2P) file-sharing systems such as BitTorrent and Gnutella, the technology world has a like-hate-love relationship. Content owners such as TV networks tentatively like them, since they make it cheaper to get high-bandwidth content like video out to viewers over the Internet. Internet service providers (ISPs) such as Comcast hate them, because P2P content clogs up their broadband networks without bringing in extra revenue. And consumers love them, because they can use the systems to download a host of songs, movies, and TV shows fast and free.
But if two venture-funded startups pairing up to make things easier for the ISPs succeed, the result might just be a like-like-love situation—or at least like-tolerate-love. That could be a big business win.
NY-based Pando Networks runs a commercial P2P network that was originally built to help people swap large e-mail attachments and has now been optimized for delivering high-definition video over the Internet. Newton, MA-based PeerApp makes caching servers that store the most-frequently-requested P2P content inside an ISP’s network, drastically reducing the amount of bandwidth sucked up by content streaming in from far-away peers. The two companies announced yesterday that they are integrating their products.
The result may be a distribution service that ISPs don’t mind hosting because it isn’t such a bandwidth hog, and that content distributors such as NBC (a Pando customer) like even more because they know it won’t rankle the ISPs as much (and because it’s cheaper than using traditional content distribution networks such as Akamai, which I wrote about yesterday).
“In many places in the media world, P2P was long deemed the enemy,” says Eliot Listman, a business development director in PeerApp’s New York office. “It was seen as this rogue methodology for distributing content without paying the content owner for the content and without paying the owners of the pipes for the infrastructure. But what Pando is selling is a P2P-based way of delivering licensed content for the same large media companies that use traditional content distribution networks like Akamai. And bringing in PeerApp couples that with the interests of the ISPs, by helping them to cache and accelerate the content.”
The enemy, in other words, is being rehabilitated. And just in time, too. It’s estimated that P2P content—that is, information that consumers are downloading from each others’ machines, using software such as BitTorrent that breaks media files into chunks and distributes them across a network of peers—accounts for as much as 40 to 70 percent of the data flowing through the ISPs’ pipes. And those pipes are being further strained by the advent of YouTube, the Apple iTunes video store, and other popular sources of streaming and downloadable video. Hui Zhang, a computer scientist at Carnegie Mellon University predicts that without new technologies to absorb the overflow, bandwidth-intensive content and applications could suck up 98 percent of the Internet’s capacity by 2009, reducing performance for everyone.
Some ISPs have attacked the problem simply by making it harder for their customers to use P2P networks, although they’ve done it secretively. In October, the Associated Press and the Electronic Frontier Foundation published the results of tests showing that Comcast—the cable TV and cable-broadband provider in Boston, the San Francisco Bay Area, and many other large metropolitan areas—was interfering with BitTorrent traffic by injecting forged packets into the data traveling between peers. Comcast denied that it blocks traffic from P2P applications, but company executives later admitted to the New York Times that the ISP does delay or “postpone” some P2P traffic to ensure quality of service for its customers.
In one way, Comcast’s attitude toward P2P traffic is understandable. Software like BitTorrent isn’t tuned to respect ISPs’ efficiency needs: if a user in Cincinnati is trying to download The Right Stuff and his BitTorrent client finds a chunk of the movie stored on the hard drive of another BitTorrent user in Norway, it will grab that chunk first, even if the guy across the street also happens to have a copy. Technologies like PeerApp’s caching servers allow ISPs to eliminate some of that long-distance traffic by storing frequently-requested files closer to those requesting them.
Listman says PeerApp has sold its servers to many ISPs in other countries, including Australia, Barbados, Colombia, Israel, Singapore, Thailand, and Venezuela. North American ISPs are still wary of P2P caching, because, by their very nature, PeerApp’s servers store chunks of content that might have been decrypted or copied illegally (as so much P2P content is). Mark Strangio, director of marketing for PeerApp, says the company’s servers store content in a way that “eliminates the possibility for legal entanglements in the unlikely event that some of that content were not appropriately protected by DRM [digital rights management, i.e. encryption].” But it’s the company’s new partnership with Pando that could really help PeerApp make its case to the domestic ISPs—or at least get them in the back door.
“The major media companies are saying ‘Okay, I buy into the economics of P2P, since the economics of traditional CDNs are so too expensive,’” says Listman. “But the ISPs have now gotten so much bad press [over P2P throttling] that the media companies are also asking, ‘When I go down that path, how I am going to make sure my content is not throttled when it hits the service providers?’ Pando is saying, ‘Let’s solve that problem by partnering with companies that do this caching.’ It’s a complementary relationship. Consumers will be happy because the kind of content they are asking for will be accelerated, and it becomes something that the service providers are pleased to provide, because it doesn’t break the bank on bandwidth.”
At least, that’s the scenario PeerApp hopes will play out. If it does, it could get home broadband subscribers one step closer to the promised land of on-demand, high-definition video over the Internet, and make PeerApp’s investors—including Boston-based Pilot House Venture Group as well as the Israel-based groups Cedar Fund and Evergreen Venture Partners—very happy. “We are dealing with a relatively new technology, and the selling process takes time,” says Strangio. “But we are very bullish on the possibilities, as are our investors. It’s a very interesting phase in the evolution of the Internet.”