Jake Loveless, CEO of Web acceleration platform provider Edgemesh, wants to speed things up across the Internet. He's spent significant time looking for ways to cut delay for traders, so he knows latency. "All it takes is one line of JavaScript code" and you can realize faster page loads using Edgemesh, he recently told me.
Edgemesh is a peer-to-peer client-side acceleration platform unlike legacy acceleration solutions that require server-side changes (Domain Name Servers, proxies). When a site becomes Edgemesh-enabled, users transparently join the Smart Mesh overlay network and are able to transfer data peer to peer. Edgemesh accomplishes this by transmitting binary data via the WebRTC DataChannel API.
According to Edgemesh, its Web acceleration platform:
- Cuts page load time by up to 60%
- Decreases bandwidth fees by up to 90%
- Enables businesses to distribute content globally without infrastructure changes
![](http://twimgs.com/nojitter/ehk/17Oct/Brunk_Edgemesh.jpg)
Jake brought these benefits into perspective by noting that "page load time correlates to sales" and "conversions are decreased by delays." which is something every savvy Google and marketing expert understands.
For the thinking behind Jake's mission to improve the Internet, read his article in ACM Queue, "Cache Me if You Can."
The Signaling System 7 telecom standard's focus on lossless failover was a big deal back in the day. Some would say improving the Internet is an a bigger deal (although some would argue the contrary). The cool thing about Edgemesh is that use of peer-to-peer (P2P) networking and a focus on cutting latency provides more benefits than just enticing a potential customer to click to buy. While P2P acceleration may not be sexy in some circles, it's pretty cool when companies can deploy the solution with no infrastructure changes that usually equate to substantial forklifting, potential disruption, and capital outlays.
What makes P2P acceleration attractive are the metrics and that the potential of P2P thriving in a neural, deep learning network environment. Yes, I believe he network will be able to act on learned data in a faster, better, cheaper way while utilizing P2P and artificial intelligence. What do you think: Is the network becoming more intelligent?
Follow Matt Brunk on Twitter!
@telecomworx