Crappy Upstream Cable Speeds Should Soon Be a Thing of the Past

With Altice and Comcast and even Charter starting to run full fiber to the prem. Does this get nationwide deployment?

There are plenty of places/subscribers where it's cost prohibitive to deploy (or replace coax with) fiber, so FDX D3.1 will help boost speeds there.

Right but as others in the rest of the comments are mentioning. Cable has been claiming to fix upload for years. My question is much like ATT made the mistake of Uverse when FiOS was rolling out. Is this really gonna be seen at any significance? Cable has always been asynchronous i get that. Yet the need for faster uploads is becoming more common. I took alot of pictures at a hiking trail the other day. My phone freaked out when i tried to email them to a family member. Better coverage with better upload fixed the issue.

So my point is if im trying to share some photos because each pic is 1-3m. Greater uploads needs are becoming a standard. With this being still a few years away. Would it not be more economical to just continue to run fiber? Even at a few miles a year for whichever provider. The glass will still be viable in another decade. The coax will see bandwidth limitations like the twisted pair of old.

Kinda seems a little futile when the glass is getting closer to everyone everyday.

The fix will be in 2 parts 1}D3.1 mid or high split(2018-19 and onward, and later(2019-20 FDX will become available)
cablecos already deploying a new split will complete the planned work, those still waiting to start MIGHT wait for FDX alone though the transition will cause the same upstream shortage during the deployment phase.
The article link in the story is 8 months old, a more current one, in the last few days shows FDX on the path to 10gig symmetric with no FTTHEven with fiber deep there are substantial costs in the FTTH drop, backhaul, and CO with no great benefit after node+0 is in place. In fact It was pointed out in the last few days how well the node+0 fdx would work with strand mount 5g micro cells as the drop AND seamless wi-fi replacement everywhere aerial plants exist

Transport long haul will continue to need to become bigger always. Regardless of end user bandwidth need aggregate bandwidth needs will continue to rise. CO back haul is not an excuse not to run fiber to the premise.

FDX while promising. Just seems like a waste. Perhaps when 5G standards are real in late 2020. Not just the current advertising buzz word. Perhaps this that will replace last mile wireline.

However still seems like cable is just repeating what ATT wasted the better part of a decade deploying. So much so they finally had to start running fiber to the premise to stay competitive.

The threat/potential of 5g is real enough to be freezing everyones long term view of the future of the drop.
AT&T uverse vs cable FTTn is not really comparable as the OLD twisted pair doesn't have near the capability of the coax drop and T's move to FTTH is because the held off so long it is their only choice where the cable plant can cover FTTH, fttn ,ftt-5g and HFC at the same time depending on what fits the market

That approach would require updating every amplifier in the outside plant twice-- first to change forward/reverse path filters (or replace the amp entirely if that's not possible), then again to swap out each amp for a mini fiber node, as required for full-duplex, so that each span of hardline coax (between amplifiers) becomes its own isolated RF domain.

It seems foolish to incur all the labor expense twice, rather than doing the FDX upgrade from the beginning, and buy thousands of new filters and/or amps that will have a service life only a couple of years... not to say they won't, but I wonder what proportion of the total cost goes toward equipment vs. labor.

They will have to. Google, AT&T, etc. have all jumped into the fiber game and CableCos will need this to stay competitive.

Not any time soon. Still waiting to see something remotely close to what Docsis 3.0 is supposed to do.

What is it not doing?

12 years after it was ratified... I get about 10% of what it's capable of downstream and 5% upstream.

Going on a limb, but i'd say for a town of 8,500 people you're lucky to have cable internet at all...

I'm rolling out 32x4 docsis 3.0 in my little town of 2000. We are an independent cable operator doing things right. Almost all of the plant is node+1 at this point, upgraded to 870mhz. Will have 32x4 and full digital video deployed by end of year. We are 8x4 right now and offer 100x25 for $70 a month standalone right now. We can deliver 50 up if everything is right but we err on the conservative side of actually delivering what we sell consistently.

well, for years So it doesn't seem to be moving anywhere.

And to date, i still don't know of a single cable modem that claims to support this nw docsis without having ass tons of problems. That retrofitted intel chipset is basically boned trying to use it, but the way it's talked about here, no cable modem maker will use anythjng else. And neither intel nor the modem makers will even address it, much less offer a solution that doesn't involve trashing and replacing the modem... With yet another with the same chipset and the same problem.

So is there even a docsis 3+ modem that doesn't have the packet drop and signal noise problems, or is it basically a lost cause at this point?

I've been able to consistently get non Puma based modems, including D3.1, so not sure what you're on about here.

Said intel modems are all i can find that one can buy, so that i don't get sacked with an additional modem rental fee.

The SB8200 is Broadcom based. Not sure about Zoom's D3.1 modem. Getting a 32x8 modem, which are generally the Puma6 based ones, is pointless on Spectrum or Comcast because 24x4 or even 16x4 will do just fine (and there are plenty of BCOM-based 16x4 modems) for the lower tiers and you can't hit the highest tier without a D3.1 box.

Both the SB8200 and MB8600 use the same BCM3390 chip.

I am on the standard tier at the house that has Spectrum and I still am rocking an 8 Channel SB6141 with the Puma 5 chipset made by Texas Instruments. It hits more than the 100 megabit downstream speed with ease. So I can confirm at least around here that even an 8 channel is fine for the standard tier.

Motorola MB8600, Arris SB8200, and Netgear CM1000 all use the Broadcom DOCSIS 3.1 chipset. Comcast's XB6 is sourced from both Arris and Technicolor..... the latter of which (the XB6-T) is powered by a Broadcom chipset.

Thanks for that. Now at least i have some ideas for when it is finally time to retire the old, enduring surfboard modem.

Broadcom has a few chipsets that are fine including 24 channel ones. If you're still at a point you only need 8 channels the Intel Puma 5 is fine because it was put out by Texas Instruments before Intel bought that division. It's the Puma 6 that is flawed. I see no sign that future chipsets will necessarily have the same flaws.

We've been hearing about higher upload speeds for years. Not just with DOCSIS 3.1, even DOCSIS 3.0 could bring upload speeds of 50-100 Mbps as it has 4 upload channels..... none of that happened, if you are lucky you get a whopping 35 Mbps upload when you get a Gigabit connection.

For me.... its seeing before believing. I'm sure it will happen some day, but I am skeptical about Gigabit in BOTH directions over cable.... I'm sure it's going to be 100 or 200 Mbps at first. And that's a maybe....

This is great news! Might put this MB8600 to real use. After seeing this news, I went ahead and upgraded from 250 -> 1000 service.

Why this won't be market ready for years!

Another way for the cable company to raise the price of the same service.

I can understand the cable Co's implementing 3.1 FDX in a few years in areas where they have fiber ISP competition, but I would surmise implementation where there is only twisted pair (e.g. DSL, VDSL, etc.) ISP competition or no competition, this won't be happening. Where's the incentive?

The El Cheapo cable companies never fully deployed the docsis 3.0 spec which REQUIRED higher grades of COAX and deeper last mile fiber than they were willing to $$ SPEND $$ on upgrades.. with VERY LITTLE competitive pressure, do you REALLY expect docsis 3.1 to be any different in this case?
You really got some gullible customer base if you think so!!!
The original docsis 1.1 spec was supposed to offer 10/10 megabits service, and after a while, cable companies decided they can service more customers by splitting the nodes and dicing the bandwidth to send more down than up while saving node spits which saved real money and allowed for the expansion of video channels which were more profitable. Now that customers are leaving for the streaming video services offered OTT for on the internet streaming from wherever.. cable companies will STILL not be chomping at the big to offer symmetrical-- only as a way to jack up rates incrementally.

This has been the promise of every DOCSIS advancement...

Rinse, lather, repeat..

Nothing has really changed in decades.. 10% of your download speed for an upload speed is still the dream on many plans.

FDX allows upstream up to 684 MHz (will grow in next version). I think that is new.