If you read the press releases or listen in on investor calls for the big cable companies over the last year, you might think that the latest cable network technology, DOCSIS 4.0, is right around the corner and will be installed soon. Cable companies have been leaving this impression to fend off competition with fiber. There are millions of new fiber passings being constructed this year where cable companies serve today, and most of the companies building fiber say that they reach at least a 30% market penetration rate within the first year after fiber reaches a neighborhood.
The reality is that it will still be a while until DOCSIS 4.0 networks make it out into neighborhoods. A recent blog from CableLabs spells this out well. This month (July 2023), CableLabs is holding the first big interoperability testing event where different manufacturers will test if their DOCSIS 4.0 equipment is interoperable with other vendors. This kind of interoperability testing is a standard step in the process of moving toward gear that is approved for manufacturing.
Per the CableLabs blog, this testing is a pre-cursor for CableLabs to be able to certify specific brands of modems. The blog describes this as the first interoperability testing event that will look to see if a cable modem can be operational when working with the latest version of DOCSIS 4.0 core equipment. This test also will check if new modems are backward compatible with earlier existing versions of DOCSIS. This is only the first of multiple interoperability tests, and later tests will go deeper into more specific functions such as interfacing with the overall network, backoffice functions, etc.
It’s normal during this kind of testing that bugs are found in the software and hardware, and it’s likely that there will still be tweaks in many of the components of the DOCSIS 4.0 network.
Only after all of the testing is done and CableLabs is happy that all components of the system are operating correctly and will work together properly can the process of certifying equipment from each vendor begin. That involves sending devices to CableLabs for extensive testing and final approval by the CableLabs Certification Board. Only then will any manufacturer put a device into mass production. Any device that doesn’t pass certification will have to be reworked, and the process started again.
It’s hard to think that it won’t be at least another year until devices start to get certified. After that will be the time needed to mass produce, distribute, and install devices. That could easily mean two years before we might see the first DOCSIS 4.0 network being installed.
With that said, this entire process has been exceedingly fast by industry standards. The DOCSIS standards was completed in early 2020. This process is far ahead of where most new technologies would be only three years after standards are completed.
The cable companies are in a huge hurry to be able to declare superfast symmetrical speeds to compete against fiber. I’m sure there has been tremendous pressure on CableLabs to speed up each step of the process. This likely meant faster than normal efforts to create breadboard chips and the components needed for equipment. For example, the normal timeline for getting a new chip designed and built can easily take 18 months. DOCSIS 4.0 chips are likely on an accelerated timeline.
Who can say how long it will take cable companies to upgrade networks to DOCSIS 4.0? They will certainly start in the markets where they think the technology makes the most market sense. It could easily take several years to make this upgrade nationwide, assuming that manufacturers will be able to keep up with the demand.
This is a subject I’m following because I firmly expect the first commercially deployed DOCSIS 4 networks to be horrible – at least from a customer experience perspective.
I’ve worked in technology labs, testing every piece of gear in the network from the customer premise gear to the core. Getting it right takes time – time the cable companies are working hard to bypass in their rush to get the new standard deployed. Unfortunately (for their customers) there is no real substitute for time. Testing these network components must be done deliberately and include every, single feature and function to ensure seamless interoperability.
But also, to ensure reliability. Cable customers have labored under poor interoperability for years. My wife and I were at the mercy of Mediacom for the first 8-9 years we moved to the Huntsville (AL) area. We’ve been on AT&T fiber for about 3 years and it’s not too bad. In fact, they just upgraded us to a Wi-Fi 6 gateway (for free). Previously, we were on a first name basis with the Mediacom field techs because they were at our house so often. Every time there was a power bump, I had to unplug the modem and DVR boxes, and then plug the modem in and wait for it to sync up and be normal. Then I had to plug in the living room DVR and wait for it to normalize. Finally, I could plug in the bedroom DVR.
Despite what you may conclude from the Charter commercials, ATT gig service is cheaper than cable (at least cheaper than Mediacom), even after the promo price dropped off after the first year. ATT has their own issues, and I’ve had to call about the bill a few times to wrangle with them on price, but by and large, compared to our Mediacom experience, ATT is miles better. I don’t know if that’s just a local issue, or because a passive fiber network is inherently more reliable than a neighborhood coax network, or all of the above.
And as much as the cable companies are hanging their hats on DOCSIS 4.0, I’m not sure it’s going to be the big win they are hoping for – at least not initially.
DOCSIS 3.1 is capable of delivering 1Gbps and higher in the upstream. The issue with deploying that to offer symmetrical speeds is not the need for 4.0, but the resistance of the cable companies to invest in upgrading and managing their cable plants to support higher split frequencies and increase the number of available upstream channels. DOCSIS 4.0 technology and silicon have been available since 2019 though it was not originally called 4.0. It has taken the pressure of significant fiber overbuilding and cable customer loss to create enough urgency to get it deployed. Cable companies have historically been slowly reactive to competitive pressure vs proactive for pushing new technology.
I have to agree with Jeff Lee here. DOCSIS 4 isn’t required to delever competing speeds. I think a lot of cable plants are running DOCSIS 3.0 right now and providers are not wanting to do a step upgrade to 3.1 with 4.0 right around the corner. 3.1 is quite good though. Down download speeds as 4.0 but 1-2G uploads… which is great!
Another hinderance is that many plants are still DOCSIS/copper fed nodes. Lashing fiber to the node and adding the optical converters is a major hurdle and moving up from a 3.0 plant to 3.1 or 4.0 is basically pointless without fiber to the node or remote CMTS with…fiber to the CMTS.
Also, lots of cable plants are heavily shared/oversold, central CMTS and splits or CMTS with fiber to the node so that 10Gx1G is way over shared.
Basically the future of DOCSIS plants is fiber in some capacity, most likely fiber to remote CMTS and coax last few hundred feet.
This brings up another concern of mine which is the infatuation with fiber to the home. DOCSIS is fantastic and cheap tech. It’s resilient, quick to install, and very fast. In many ways, DOCSIS is superior to fiber because of it’s resilience. Fiber has exactly one benefit which is distance. It’s no ‘easier’ to shove bits down a fiber than bits down copper short distances and the side effect of fiber’s long distance abilities is more crossing, more poles, more chance for a single downed pole or backhoe hit to take out everyone because fiber networks have little resilience.
Most cable users are NOT limited by the backhaul (the actual docsis part). Most have more issues in the home network, generally Wi-Fi related. Most consumers don’t understand wi-fi at all, and there are lots of limitations based on distance, generation of Wi-Fi, number of walls the signal is going through etc. They try to sit in a room two stories from the access point, going through three walls, on a cheap $49 access point and run video or gaming. Its not going to work that well. There are some fundamental limitations introduce by physics, old technology, interference with your neighbors and cheap implementations. DOCSIS 4.0, and for that matter Wi-Fi 7, is going to have significant impacts, in terms of speeds, for a limited set of customers. A few mesh nodes and some new client devices will help a lot more people and can be done much more cost effectively.