Blog Post

100 Mbps DSL is Here & 800 Mbps is Around the Corner

Stay on Top of Enterprise Technology Trends

Get updates impacting your industry from our GigaOm Research Community
Join the Community!

It is mind boggling to think that copper, thanks to new generation DSL technologies is staying competitive with fiber and cable broadband. Today, a new breakthrough shows that it will only be a matter of time before DSL broadband crosses the 800 Mbps threshold. And while we wait for that massive speed bump, we are beginning to see the commercial availability of DSL that can deliver 100 Mbps.

These recent upgrades in the DSL speeds are coming at a handy time – DSL has started to lose market momentum, and carriers are looking for ways to balance their exploding capital expenditure requirements. While fiber networks are better in the long run, most phone companies need to squeeze out more from their copper networks without losing too much ground to cable broadband rivals. Why? Because they have to shift their capital expenditure dollars to beefing up wireless networks, which themselves are growing through an explosive growth.

Ikanos (s IKAN), a maker of broadband chips today introduced a new technology, NodeScale Vectoring, DSL access technology that allows connections at 100 Mbps and higher, something which has not been possible on many of the phone company networks. According to the chipmaker, the cost of deploying this technology is about a tenth of the cost of building a fiber to the home network.

The NodeScale technology allows carriers to eliminate the crosstalk that occurs on copper pairs when offering very high-speed Internet. The cross talk introduces noise in the network, which in turn limits the line quality and thus reduces the performance of the network. Typically, to handle crosstalk issues, one needs gigabytes of memory. There are two ways of handling cross talk. NodeScale essentially tames cross talk at the DSLAM level as opposed to line card vectoring which treats every line card as a separate crosstalk domain.

Ikanos claims its NodeScale Vectoring technology cancels noise efficiently, and ZTE Corporation will demonstrate the first DSLAM employing the technology.  The technology was developed in-house, but Ikanos also licensed the dynamic DSL technology developed by DSL pioneer John Cioffi’s ASSIA.

Back in 2006, Professor Cioffi (of Stanford University) told me that it would be possible to hit Gigabit speeds over DSL. We are inching pretty close to that. We have written about many experiments which have pushed DSL speeds to over 300 Mbps in lab conditions. Huawei, the Chinese telecom equipment maker recently announced that it has tested speeds of up to 700 Mbps.

Today, Nokia Siemens Networks took that even further, announcing that it has tested a technology that could boost the data-carrying capacity of standard copper wires to 825 Mbps over a distance of 400 meters of bonded copper lines and 750 Mbps over a distance of 500 meters. If it is made commercial, it would allow the carriers to eek out more from their copper infrastructure.

NSN does this by the creation of phantom (or virtual) channels that “supplement the two physical wires that are the standard configuration for copper transmission lines.” The approach is called Phantom DSL and can boost bandwidth by between 50 percent to 75 percent over the existing bonded copper lines. NSN hopes to make it part of its DSLAM products. Alcatel-Lucent’s (s ALU) Bell Labs came up with the Phantom DSL technology and announced it back in April 2010.

Related GigaOM Pro Content (sub req’d):

10 Responses to “100 Mbps DSL is Here & 800 Mbps is Around the Corner”

  1. Hah I’ll believe it when I see it delivered consistently. At home I’m supposed to get 6 Mbit/sec from AT&T and at night it drops down to about 1.3 so I’m guessing many ISP’s won’t have capacity internally to handle such speeds. Of course they’ll probably add in usage caps to make it appear they have capacity, but like I said, I’ll believe it when I see it.

  2. Last mile bandwidth (in Telco networks) isn’t the main area where capacity improvements are needed to improve end user experience. 30Mbps to the home would allow multiple HD live streams + support downloads and other activities.

    The congestion points happen further back in the network (e.g. DSLAM/OLT uplink, aggregation router, router cores, peering points). Increasing the b/w & reducing the overbooking factor in these areas would could make a 15Mbps service seem uber fast for most consumer applications.

    Last mile b/w needs to increase of course to enable applications that will come (3DTV, holograms, …), but that isn’t the bottleneck today.

  3. I wonder how much of this information is part of a infomercial that Om agreed to write and got paid for and how much of it is purely to discuss the state of DSL.
    This stinks more of an infomercial masquerading as a nice blog post.
    Everything that has been discovered about DSL states the limitations for throughput whether it is signal noise or the distance you can sustain, yet people continue to hype it and try to continue to sustain themselves and their stock value.
    I believe the most effective solutions for broadband access lie in Wireless broadband and the big bets made by telco’s in FTTH like Fios or FTTC like Uverse are in the long run not very cost effective when they compete with wireless broadband, will they shut them down is to be seen. Will the U.S. FCC allow Municipalities to offer their own wireless broadband and or will we need a tea party movement to challenge the status quo that we have of high priced broadband access. DSL, Cable and Fiber all are expensive and we have oligopolies operating not free competitions.
    Lets see some real blog post by Om on this rather than a infomercial….

  4. chrisconder

    Its all a con to protect the copper cabal for longer. They may even have to lay more copper to make it work, which is criminal when they could be laying fibre. Bonding obsolete copper pairs is old tech. Bring on the light.

  5. Dave Burstein

    Unfortunately, the 100 meg only works over very short distances and requires replacing line cards or DSLAMs. That’s unlikely, so this won’t help most of us. Related improvements in noise reduction should be good for a 15% to 50% improvement over what you have. Every carrier has the technology to double your speed using two lines, but as far as I know in California is the only one offering that as a commercial product.