The real question though is will the new "modem mode" firmware also have fixes for all the Superhub bugs. It may be that the memory leak VM have identified was responsible for all of them but there could be other errors too. Most of the people interested enough to sign up for the trial are going to only be interested in using their own routers so how much testing will the standard hub actually be getting?

Many Superhub complaints related to WiFi range - I doubt that any firmware change will improve the range and the feature set will remain very poor compared to even a £20 router. Its only redeeming feature is the gigabit switch.

I did see that VM may well be adding downstream channels and if they do that I'll probbly have to get my Superhub re-authorised in place of the VMNG300 I got from the CEO office when it became obvious VM weren't able to fix the Superhub bugs in any sensible timescale and were never going to improve the feature set.

its getting testing, they even did R29T4/R30 to specifically try and fix some wireless issues.

In terms of the range and such that is in my view down to poor hardware design. For me the internal antennae was a bad idea for a start. People are reporting much more stable wireless usage now tho.

Also i dont see any new features on R29 so yeah it looks like only modem mode is new, but that will satisfy the majority of geeks. The superhub is a router for newbies basically, probably what VM want it to be.

The extra downstream channels wont stop a vmng300 working, it will just not be able to use them so its more likely to see congestion in that scenario. Since my vmng300 consistently has much better jitter than my superhub it would need to be quite severe downstream congestion for me to do the swap as jitter is important to me. So far its just 1 extra downstream channel in 'some' areas so an extra 20% downstream capacity.

Edited by Chrysalis (Sun 04-Sep-11 13:40:25)

I guess time will tell. I only saw issues myself when using wireless and I would have accepted the poor feature set if it had been stable and had better range.

5 channels instead of 4 is 25% more and considering how small the pipe is to start with I'd say it's worthwhile. Have you seen any numbers for the contention ratio VM use? With a tiny 200Mbps pipe at the local level I'm frankly amazed they can support 100Mbps at all.

BTW I thought most jitter issues on cable were due to upstream? Do you know why you have less jitter with the modem than the hub? Do others report that too?

In reply to a post by kwikbreaks:

I guess time will tell. I only saw issues myself when using wireless and I would have accepted the poor feature set if it had been stable and had better range.

5 channels instead of 4 is 25% more and considering how small the pipe is to start with I'd say it's worthwhile. Have you seen any numbers for the contention ratio VM use? With a tiny 200Mbps pipe at the local level I'm frankly amazed they can support 100Mbps at all.

BTW I thought most jitter issues on cable were due to upstream? Do you know why you have less jitter with the modem than the hub? Do others report that too?

I am amazed also.

Incidently I have tried 100mbit and the results werent good so I downgraded again.

On 30mbit I get full speed outside of peak pretty much all the time, during peak its started to get flaky, I upped to 100mbit to see if I could give them justification to upgrade capacity as 30mbit slowing down to 25mbit or so isnt particurly serious. I was expecting full speed off peak and 25mbit on peak. The on peak was right I got about 25mbit on the 100mbit service but off peak it only hit about 60mbit. With that performance VM said utilisation was high but not high enough to justify a capacity upgrade so they consider 25% peak and 60% off peak good enough for my area on their top tier product. My area does not have 5 channels and also only has 2 upstreams whilst some other areas have 3 upstreams. Also upstream did hit full speed off peak but down to about 2mbit/sec on peak.

Jitter is primarily caused by upstream congestion of which upstream congestion is the majority of congestion VM has.

My 2 theories are.

1 - the vmng300s have a higher priority given they were originally assigned to the old top tier product that would have some logic.
2 - the way the modem works and is configured probably with a smaller upload buffer so as such reduces latency/jitter.

I have had another person get the same with me on jitter however if the area is good enough with utilisation I think there is no difference between the 2.

Whilst an extra 25% is something I just feel my connection slowing down to 25mbit doesnt feel much more laggy when using it yet when I have the higher jitter web browsing is noticebly less snappy, using ssh is awkrawd and it affects my nephews gaming as he occasionally plays online on the xbox when here.

In terms of numbers ignition reported 200 or so modems on his shared upstream channel on the VM forums. I think its becoming more evident VM dont care about contention ratio anymore the fact congestion is noticeble even with 3 forms of traffic management active. I suspect their policy for upgrading is when a few complain give them all discounts, when enough complain then do an upgrade which itself usually takes 6 months.

Edited by Chrysalis (Sun 04-Sep-11 20:50:54)

just to pick out a couple of things chris, hope you dont mind

Ignition says about 260 modems on his upstream. Thats on non-uplifted upload speeds

and upstream bonding is set to be trialled early next year, so that may have an impact on jitter

In reply to a post by craigj2k10:

just to pick out a couple of things chris, hope you dont mind

Ignition says about 260 modems on his upstream. Thats on non-uplifted upload speeds

and upstream bonding is set to be trialled early next year, so that may have an impact on jitter

the bonding will have an impact however will it be extra capacity (new channels) or sharing of existing channels. VM really need to get QAM64 working on the upstreams, they miles behind comcast.

In typical VM style they will probably trial the bonding in low utilisation areas (the same areas that got the uplifted speeds early) and then rollout the result in higher utilised areas that need a different capacity plan.