With the release today of its next generation of Gfast chipsets, Sckipio delivers up to 2Gbit/s speeds in both downstream and upstream directions, supports up to four times as many subscribers sharing one binder and further enhances speeds through the use of 212MHz bonding for CPE and DPU.
The Sckipio SCK-23000 family has four times the vectoring capability, twice the bonding speeds and supports coordinated DTA (cDTA) for symmetric-like performance in downstream and upstream, said David Baum, Sckipio CEO.
The chipset includes cross-DPU vectoring without the need for external vectoring, allowing smaller DPUs to connect to increase vectoring port density, said Michael Weissman, co-founder of Sckipio, in an interview with UBB2020. This allows telcos to "pay as you grow," he said.
"You can start with a small box, an 8-port box, and as you grow you can add another 8-port box, then another 8-port box," he said. "Or, if you're in a larger environment, maybe you start with a 16-port box, then add another 16-port box and a third 16-port box. You can get up to 48 ports when you start with a 16-port. This means you have one third of the capital expense upfront and your return to profitability is much shorter because you only have to build the first box. You don't have to build all the boxes before you're profitable."
Telcos also can use one SKU yet create up to eight designs for different use cases, added Weissman. From an opex perspective, a service provider can use one box to validate, certify, deploy and train field installation technicians and so forth, he said. Finally, this approach enables CSPs to reverse power-feed because they can use smaller CPEs to keep the DPU live, said Weissman.
"At some point you can't make the box too big because you can't make the CPE too expensive and in some cases there may be regulatory limits on how much power you can inject before it gets too unhealthy. You can't do injected reverse power-feed on a 48-port box. It requires too much power," he said. "With cross-DPU vectoring, I can use a 16-port box, have a single subscriber, power up that 16-port box; I can add another 16-port box, have another single subscriber and power up that box. I can add another 16-port box. I'm at 48 ports but I'm only having to power up 16-port boxes, which is very do-able with the current limits from ETSI, the capabilities of the system and the low power requirements of our chips. So what that means is you can actually get to 48-point vectoring at 212 in a way that literally was not possible before."
The chipset is designed to help operators further compete against cable companies' rollouts of DOCSIS 3.1, said Weissman. In part, it does this by featuring cDTA support beyond 1 Gbit/s in upstream and downstream. (See DT, ADTRAN Lab Test Super-Speedy G.fast .)
This allows service providers to deliver gigabit service to consumers without installing costly, time-consuming fiber-to-the-home (FTTH). And, if CSPs combine cDTA with bonding, resulting speeds hit 2 Gbit/s, said Weissman.
Some service providers will offer bonding-as-a-service (in fact, six months ago Windstream announced a partnership with Sckipio to provide this offering), he said. Others will use bonding to bring high-speed Internet to subscribers plagued by poor bandwidth due to something in the their home or a regional problem, Weissman said.
Speedier Gfast also empowers service providers to resolve the urban digital divide, he said. Although Gfast might not answer rural broadband problems due to a lack of copper and long distances, cities are prime candidates for next-generation Gfast, Weissman said. (See Disconnected in the City.)
"In the urban environment, there is also a digital divide between the rich and the poor. Gfast is the best solution," he claimed.
— Alison Diana, Editor, UBB2020. Follow us on Twitter @UBB2020 or @alisoncdiana.