|
[Connectivity] Is OFDM used even if you have sub-gig package?Just wondering if the DOCSIS 3.1 OFDM is actually used if you do not have gigabit service. On a Comcast forum thread, I said it should be, but somebody said he had been told it's only used for gig speed. Anyone have a definitive answer? » forums.xfinity.com/t5/Yo ··· /3414493 |
|
DocDrewHow can I help? Premium Member join:2009-01-28 SoCal 1 edit
12 recommendations |
DocDrew
Premium Member
2021-Feb-14 3:42 pm
Yes it's used on every modem that can use it. It's added capacity is bonded to the QAM channel capacity to make a bigger pipe. It doesn't benefit the provider or the customer not to use it. |
|
3 recommendations |
to andyross
Ofdm is used with any tier My parents are on the 25mbit tier and get 31 downstream channels and a 96mhz ofdm channel We are all still waiting for ofdma |
|
ka2107 join:2021-02-03 Saint Charles, IL ARRIS S33 Dynalink DL-WRX36 (Software) OpenWRT
1 recommendation |
to andyross
I have Extreme Plan 400/20 Mbps in Chicago Suburb. I am using Arris SB8200 right now and I have DOCSIS 3.1 Channel enabled for Downstream. I have a friend living in a different Chicago Suburb using XB6 with Performance Plan 100/5 Mbps and he also has DOCSIS 3.1 enabled in the Downstream. Neither of us have DOCSIS 3.1 enabled in the Upstream. Currently both of us have only 5 DOCSIS 3.0 Channels in the Upstream. |
|
srtdodge05 Premium Member join:2011-10-16 Ypsilanti, MI |
to andyross
It might use it but does it need it no. |
|
11 recommendations |
Anon163f8 to DocDrew
Anon
2021-Feb-14 7:01 pm
to DocDrew
said by DocDrew:... It doesn't benefit the provider or the customer not to use it. (and while you are aware of this), I think just as importantly it should be noted that it *does* benefit the provider and the consumer to use it whenever possible, as OFDM is spectrally more efficient and more tolerant of signal issues such that not only does it free up capacity for those that (for whatever reason) do not have OFDM capable modems (potentially reducing "The Internet is slow at night/day/all the time" complaints for everyone), it can also reduce the impacts of certain disturbers (as OFDM can dynamically choose to not use impaired sub-channels). A win/win for the provider and the customer (well, the provider does have some additional fees, but still) is a hard thing to pass up. |
|
4 recommendations |
to andyross
Yes, and it appears to be prioritized over using QAM256.
I have 300 down, 10 up on a CM1200 and most codewords are passing through my ODFM channel. |
|
5 recommendations |
Anonf49bb
Anon
2021-Feb-15 12:13 pm
Robbie is 100% correct. If you have a D3.1 modem and there is an OFDM channel available, the modem will use bandwidth on that channel first.
The reason why they'd want any modem that is capable of using it to use it is because DOCSIS3.1 is backwards compatible. You either have a modem that only supports legacy, or you have a modem that supports legacy and modern concurrently. So if 100 people on slow plans use OFDM then people on gig can still get plenty of bandwidth from legacy channels since those have been freed up. Where it would be an issue is if the modems weren't backwards compatible, then you could run into that issue where everyone is attempting to consume all of the OFDM bandwidth but the legacy is going largely unused. |
|
Anonf49bb
4 recommendations |
to srtdodge05
You can't really make an assessment whether or not non gig tiers need to use OFDM as a blanket statement. It's simply more bandwidth that is available to the user. The plan only dictates the amount of bandwidth you can use, not how much you're actually using. A dozen people on 100 meg performance plans could consume as much bandwidth as a gig user, and I doubt it takes even that many to equal the bandwidth of a gig user. So if your node is largely filled with D3 modems and all of the channels are saturated, then even a 25 meg user is going to benefit because their modem isn't fighting for bandwidth with all of the other users. This would be the primary reason why it's the priority channel for any modem capable, since it offloads that data from D3 for modems that can't take advantage of D3.1. If you offload ten 100 meg users to OFDM then it allows someone who's still D3 but say a 400 meg user to continue using their connection without interruption. |
|
Technicolor CGM4981 Ubiquiti EdgeRouter Lite Motorola MB8600
1 recommendation |
to Anon163f8
said by Anon163f8 :A win/win for the provider and the customer (well, the provider does have some additional fees, but still) is a hard thing to pass up. I'm surprised that Comcast doesn't try to move people to docsis 3.1 modems with an incentive. If they offered even a small incentive like $2 - $3 per month credit for a 3.1 modem, it would cover much of the cost difference of the better modem over the lifetime of the modem and make it a losing proposition to buy anything less. That would allow them to move to OFDM upstreams sooner and make mid splits much more desirable. |
|
5 recommendations |
First they need to stop renting out gateways with those horrible Intel Puma CPUs. |
|
5 recommendations |
to Anonf49bb
To add: It is the CMTS that does all the work. It loads the channels with data in the best possible way. The modem just sits back and retrieves data from the different channels. |
|
mixdup join:2003-06-28 Alpharetta, GA
1 recommendation |
to madbavarian2
said by madbavarian2:said by Anon163f8 :A win/win for the provider and the customer (well, the provider does have some additional fees, but still) is a hard thing to pass up. I'm surprised that Comcast doesn't try to move people to docsis 3.1 modems with an incentive. If they offered even a small incentive like $2 - $3 per month credit for a 3.1 modem, it would cover much of the cost difference of the better modem over the lifetime of the modem and make it a losing proposition to buy anything less. That would allow them to move to OFDM upstreams sooner and make mid splits much more desirable. Getting a few more people over to 3.1 modems isn't really worth it. There are better ways to get bigger bang for the buck on the downstream side. And whether people are on 3.1 modems or not has zero impact on moving to mid-split for the upstream. Sure, it would allow them to use OFDM or other higher-order modulation, but the bulk of the work to get to mid-split has nothing to do with modems |
|
1 recommendation |
Anon33725
Anon
2021-Feb-15 6:07 pm
said by mixdup:.... but the bulk of the work to get to mid-split has nothing to do with modems Including, no doubt, figuring out how to get GRF-54-85's (or equivalent) available/deployed to customers for self install in mixed residential environments (I believe the issue was mentioned in the conference presentation(s), but I do not recall seeing a firm plan). |
|
mixdup join:2003-06-28 Alpharetta, GA
1 recommendation |
mixdup
Member
2021-Feb-15 6:09 pm
said by Anon33725 :said by mixdup:.... but the bulk of the work to get to mid-split has nothing to do with modems Including, no doubt, figuring out how to get GRF-54-85's (or equivalent) available/deployed to customers for self install in mixed residential environments (I believe the issue was mentioned in the conference presentation(s), but I do not recall seeing a firm plan). or replacing set tops entirely |
|
1 recommendation |
to mixdup
The simplest way to get more bang is doing node splits. Comcast has already addressed that by doing Node + 0 wherever it makes sense. That is a direct impact to downstream, upstream, and CNR for the node. If there is anything to hang your hat on is that Comcast realized this earlier on and it makes tackling all of the other issues a bit easier. Smaller node sizes means less customers to deal with at once to get a mid-split online. |
|
mixdup join:2003-06-28 Alpharetta, GA
4 recommendations |
mixdup
Member
2021-Feb-15 6:54 pm
said by bman212121:The simplest way to get more bang is doing node splits. Comcast has already addressed that by doing Node + 0 wherever it makes sense. That is a direct impact to downstream, upstream, and CNR for the node. If there is anything to hang your hat on is that Comcast realized this earlier on and it makes tackling all of the other issues a bit easier. Smaller node sizes means less customers to deal with at once to get a mid-split online. They've been talking about mid-split and other strategies to increase upstream for a decade now. I'm done giving them credit |
|
|
4 recommendations |
said by mixdup:They've been talking about mid-split and other strategies to increase upstream for a decade now. I'm done giving them credit It is not just talk. They are doing a lot of work behind the scenes. It is only a matter of time before they release some exciting news. |
|
1 recommendation |
to mixdup
I'm not going to disagree with you at all on that one. After having the same upload for a decade they better lay down some plans soon. I've seen better speeds off non MM Wave 5G which costs less and doesn't have a data cap. They've fallen behind some cable companies which they were years ahead of in implementation. The sad reality is that they are still doing better than some others that are out there. |
|
ARRIS S33
4 recommendations |
to Franken
said by Franken:said by mixdup:They've been talking about mid-split and other strategies to increase upstream for a decade now. I'm done giving them credit It is not just talk. They are doing a lot of work behind the scenes. It is only a matter of time before they release some exciting news. Let's do all this work to make mid split work then add a 6th channel at 10 mhz They've been doing it since 2015 and all they have to show for it is a test home here and there, and not to forget about the 1 test home high split with 1.25gb symmetrical connection. 1 home. Itll be another 10 years since announcing the 1 home for it to become 1 neighborhood |
|
mixdup join:2003-06-28 Alpharetta, GA
3 recommendations |
to Franken
said by Franken:said by mixdup:They've been talking about mid-split and other strategies to increase upstream for a decade now. I'm done giving them credit It is not just talk. They are doing a lot of work behind the scenes. It is only a matter of time before they release some exciting news. They have been teasing "exciting news" for a decade. I'm not doubting that they're doing a lot of work and spending a lot of money. They have been saying they were doing that for 10 years now and we're where we've always been. My point still stands. |
|
3 edits |
to videomatic3
said by videomatic3:Let's do all this work to make mid split work then add a 6th channel at 10 mhz About that. They are not mid-split everywhere, and different care has to be taken when they go live above 42 MHz. Two very different cases. According to their expo presentation, their fiber deep areas (mid-split ready) have been doing very well during the pandemic, so the 6th channel may not even show up here. |
|
1 recommendation |
They’re not midsplit anywhere. |
|
mackey Premium Member join:2007-08-20
3 recommendations |
to Franken
said by Franken:According to their expo presentation, their fiber deep areas (mid-split ready) have been doing very well during the pandemic Not giving people shit for upload bandwidth + small node sizes = no bandwidth problems? You don't say. said by Franken:It is not just talk. They are doing a lot of work behind the scenes. It is only a matter of time before they release some exciting news. Please, these claims were being made for over a decade now, I'll believe it when I see it getting widely deployed. No matter how much "behind the scenes" work they are doing (lol, you don't actually believe that bullshit do you?), it will still take 10+ years to perform the physical work on the the equipment in the field to get it to any significant portion of their footprint. |
|
ARRIS S33
3 recommendations |
Lol, I knew I wouldn't have to reply, someone would.. We see charter and cox deploying better upload capacity more actively than comcast, they actually have odfma or 8 upstream channels, this to me looks like they are trying harder than comcast. Comcast claims to be the leader but we have yet to see. This feels no different than att deploying 6mbit to their fiber users when 24mbit was available to dsl users, so people have an "equal experience" I'll be happy if comcast proves us wrong, but even when they deployed gigabit it took 2 years to get "everywhere" |
|
|
to mackey
said by mackey:Please, these claims were being made for over a decade now, I'll believe it when I see it getting widely deployed. No matter how much "behind the scenes" work they are doing (lol, you don't actually believe that bullshit do you?), it will still take 10+ years to perform the physical work on the the equipment in the field to get it to any significant portion of their footprint. They began their fiber deep / mid-split project in 2015. They teased about a 100 Mbps upgrade in a presentation in 2017. Yes, I believe they are doing the work that they say they do. |
|
ARRIS S33
8 recommendations |
Remmeber this one in 2013? Good news is we are almost 10 years from that one so maybe we get 75mbit up in 2023 100mbps in 2025, and finally symmetrical in 2030 |
|
mixdup join:2003-06-28 Alpharetta, GA
2 recommendations |
to Franken
said by Franken:said by mackey:Please, these claims were being made for over a decade now, I'll believe it when I see it getting widely deployed. No matter how much "behind the scenes" work they are doing (lol, you don't actually believe that bullshit do you?), it will still take 10+ years to perform the physical work on the the equipment in the field to get it to any significant portion of their footprint. They began their fiber deep / mid-split project in 2015. They teased about a 100 Mbps upgrade in a presentation in 2017. Yes, I believe they are doing the work that they say they do. Not sure if you've looked at a calendar, but 2015 was 6 years ago. A lot of work with nothing to really show for it |
|
3 recommendations |
said by mixdup:Not sure if you've looked at a calendar, but 2015 was 6 years ago. A lot of work with nothing to really show for it That is up to the decision makers at Comcast. Comcast is in the process of upgrading their entire infrastructure + deploying fiber deep nodes. I call that a lot of work behind the scenes. |
|
2 recommendations |
They must not be very confident in the work they have accomplished so far, otherwise we would see them deploying 8 channel upload configs. Nobody is questioning the scale of the work required to accomplish the midsplit. The problem is that they are years behind schedule. Here on the peninsula, it’s mostly overhead lines, so I have some visibility into their progress with node 0. To call it 1% complete would require a margin of error of 1%.
I’ve said it once and I’ll say it again: if I had a dollar for every trunk amp on the peninsula, I could afford to buy a house here. |
|