802.11ax vs. 10GigE: Which will come first in the home?

[Gryz]

It's all about the price. A quick look revealed that most 10Gbps-cards cost 300-400 euros. Some have 2 ports. But still, 300-400 euros is very expensive for home equipment.

When that price drops down to 50 euros, a lot of enthousiasts will start buying 10Gbps cards. And high-end NASes and other stuff will have 10 Gbps (standard, or as an option). And once that wave starts rolling, costs will drop even further. And then all mothetboards will have 10Gbps. It might take 5-10 years, but we will get there. There will be progress. We're not at the end of the line yet.

WiFi is completely different. We'll see bigger marketing numbers on cardboard boxes. But real-life performance will just suck. As it does today. There will be WiFi of course. You need it for your handheld devices. But anyone doing anything half-serious will want a cable with 10Gbps. Just takes time.

 

[ksec]

Let put a few things in our mind,

802.11ax is basically what the wireless world learn about LTE-A ( OFDMA ) and incorporate some of their wisdom in to WiFI.

802.11ax promise to bring 4x REAL world speed improvement compared to 802.11ac. Our real world average speed of 802.11ac range from 150mbps to 500Mbps for dual stream, 4x that we get 600Mbps to 2Gbps.

802.11ax also works in 2.4Ghz as well, so you get the extra bandwidth on 2.4Ghz

MU-MIMO is no longer an option features, although much like 802.11ac this is download only. Upload MU-MIMO will be an 802.11ax features but still optional ( Likely an Wave 2 thing, along with 8 concurrent MU-MIMO client )

If we continue the current marketing numbers game, 8x8 802.11ax will likely have 20,000+ number, i.e 20Gbps.

-------------

Much like EUG said, NAS is now being bottlenecked by 1Gbps Ethernet. When you consider a multi user environment all with 1Gbps connection, you can see the same 1Gbps NAS would be the bottleneck for all file access. And if you are in SME where your NAS will likely be Raid 5+, the transfer speed will easily top 100Mb/s. And with SSD getting cheaper, we are increasing using it as Hot file cache.

------------

In properly 90% of the case, 1Gbps Ethernet is enough, as most have said. But in a Post-PC era, where most consumer will be Mobile, Tablet, Laptop connecting via Wi-Fi, these cable connection will quickly become the backbone for us. Even current top of the line 802.11ac router can saturate the 1Gbps port, that is why the industry introduce NBase-T.

NBase-T is based on 10Gbps 10Base-T tech, allowing 2.5/5Gbps speed on good old CAT5e / CAT6 cable. Most home users, with their much short length of installed cable at home should be able to get 5Gbps speed. I am not sure if NBase-T actually include 10Gbps inside the spec.

------------

So basically for everywhere that cables continues to live on, there is definitely a huge need to increase speed beyond 1Gbps. Especially for router and or wireless AP.

 

[Eug]

Thx, I wasn't aware of N Base T. Ironically I always thought it would beneficial to have some sort intermediate technology that didn't require such restrictive specifications as 10 Gig E, suitable for home use and SOHO applications.

5 Gigabit would be perfect for me actually, and would last me a long, long time. BTW when I wired my home for Ethernet many, many years ago I only used my leftover CAT5e and then CAT6. I never bothered with CAT6a because the cables are so much thicker and because at the time it was very expensive. All my runs are under 30 m anyway, which would with work CAT6 for 10 Gig E, but several of my runs are still CAT5e only.

If N Base T means faster time to market and lower cost parts for the home and SOHO markets, then I'd be all for that. Furthermore, given that it shares technology with 10 Gig E, this could mean manufacturers could create hardware that auto-adjusts speeds from 10 Gigabit to 5 Gigabit to 2.5 Gigabit, instead of dropping all the way down to 1 Gigabit. This would be HUGE for us consumers.

 

[ksec]

Thx, I wasn't aware of N Base T. Ironically I always thought it would beneficial to have some sort intermediate technology that didn't require such restrictive specifications as 10 Gig E, suitable for home use and SOHO applications.

5 Gigabit would be perfect for me actually, and would last me a long, long time. BTW when I wired my home for Ethernet many, many years ago I only used my leftover CAT5e and then CAT6. I never bothered with CAT6a because the cables are so much thicker and because at the time it was very expensive. All my runs are under 30 m anyway, which would with work CAT6 for 10 Gig E, but several of my runs are still CAT5e only.

If N Base T means faster time to market and lower cost parts for the home and SOHO markets, then I'd be all for that. Furthermore, given that it shares technology with 10 Gig E, this could mean manufacturers could create hardware that auto-adjusts speeds from 10 Gigabit to 5 Gigabit to 2.5 Gigabit, instead of dropping all the way down to 1 Gigabit. This would be HUGE for us consumers.

That was exactly what i asked in the other thread. It would have been nice if NBase-T controller allows 10Gbps.
The way it works now, 10Gbps uses 1 lane, and if you want 40Gbps you use 4 lane, and 10 if you want 100Gbps. The newer 25G up the per lane to 25Gbps, which means you only need 4 lane to get 100Gbps.

Edit: I found the answer, NBase-T will be basically a dumped down version of 10Gbps. So yes, if your cable allows you will get 10Gbps.

 

[rchunter]

I imagine by Skylake E/EP we should start seeing some motherboards with 10GbE integrated onboard....Once that happens hopefully 10GbE switches will start dropping in price rapidly.

 

[Engineer]

I imagine by Skylake E/EP we should start seeing some motherboards with 10GbE integrated onboard....Once that happens hopefully 10GbE switches will start dropping in price rapidly.

Throw in the fact that SoC chips like Xeon D (1518, 1540, more) have dual 10GbE built into the SoC and I'm sure that other, consumer class SoC will follow, helps this idea also.

The Nbase-T looks interesting. I wired my home with Cat6 this past summer and would be nice to go above 1Gbps later on at an economical cost (new NIC and new switches).

 

[ksec]

Throw in the fact that SoC chips like Xeon D (1518, 1540, more) have dual 10GbE built into the SoC and I'm sure that other, consumer class SoC will follow, helps this idea also.

The Nbase-T looks interesting. I wired my home with Cat6 this past summer and would be nice to go above 1Gbps later on at an economical cost (new NIC and new switches).

The problem is I dont see how 10gbps component will get any cheaper this year, or even next year. I am hoping together with USB 3.1 ( 10Gbps ) could leverage some economy of scale.

 

[Engineer]

The problem is I dont see how 10gbps component will get any cheaper this year, or even next year. I am hoping together with USB 3.1 ( 10Gbps ) could leverage some economy of scale.

It will get cheaper at some point. Just ask my first hard drive: 20MB (yes, MB) - MFM drive with an interleave of 4 to 1 - $379.99 plus $9.99 shipping. :biggrin:

 

[StrangerGuy]

It will get cheaper at some point. Just ask my first hard drive: 20MB (yes, MB) - MFM drive with an interleave of 4 to 1 - $379.99 plus $9.99 shipping. :biggrin:

You are a comparing a time when consumer computing storage is so critically limited in its infancy versus a now where wired GbE is more than enough bandwidth for a vast, vast majority of home users. I can say with certainty that hardly anybody is going to gives a rat's ass about bandwidth as long as their streaming videos works properly which ironically is going to be much easier with H265/VP10 that cuts bitrate by half for the same quality.

 

[rchunter]

It will get cheaper at some point. Just ask my first hard drive: 20MB (yes, MB) - MFM drive with an interleave of 4 to 1 - $379.99 plus $9.99 shipping. :biggrin:

lol, I had a 20MB HDD too. Apple hard disk 20 on my old mac plus cost me like $1500.

 

[rchunter]

You are a comparing a time when consumer computing storage is so critically limited in its infancy versus a now where wired GbE is more than enough bandwidth for a vast, vast majority of home users. I can say with certainty that hardly anybody is going to gives a rat's ass about bandwidth as long as their streaming videos works properly which ironically is going to be much easier with H265/VP10 that cuts bitrate by half for the same quality.

What about when people start commonly ripping bluray 4K movies and putting it on their file servers? I know I for one am looking forward to higher bandwidth to and from my servers. As soon as I can buy a switch for a reasonable cost I plan on upgrading to 10GbE.

 

[Engineer]

What about when people start commonly ripping bluray 4K movies and putting it on their file servers? I know I for one am looking forward to higher bandwidth to and from my servers. As soon as I can buy a switch for a reasonable cost I plan on upgrading to 10GbE.

Not just ripping but streaming them online. I remember at the time of my 20Meg HD that I said that I would never fill this thing up. The comparisons then are valid now. There's no difference. Once mult-gig connections start hitting homes at a high level, faster stuff will be released or the current stuff will come down in price for the home consumer. It may not happen this year....it may not happen next year but it will happen.

As for the compression comment, that may be true. Of course, compression was released just after I bought the 20Meg drive and it was going to extend the storage and life of my drive even more than it was already (remember, I was never going to fill it up and now it could be compressed on the fly making it even better).

4k streaming just starting up and 8k streaming (or more) will be next (not sure why though....how much more detail do we really need to watch video / tv?)

 

[MtnMan]

For home, it doesn't really matter. 10gbE will likely never be necessary outside extreme enthusiast markets. If you're really that impatient, then you can pay the extra.

People used to say that 802-11b was all anyone needed, because very few people had broadband even approaching the 11Mb speed of -b

 

[StrangerGuy]

What about when people start commonly ripping bluray 4K movies and putting it on their file servers? I know I for one am looking forward to higher bandwidth to and from my servers. As soon as I can buy a switch for a reasonable cost I plan on upgrading to 10GbE.

Yet another tired "niche exist so it isn't niche" argument. Forget 4K, what % of people out there actually even rip Blurays compared to stream videos from the Internet?

If I get $1 for everytime I see that on PC tech forums I will be rich by now.

 

[XavierMace]

Yet another tired "niche exist so it isn't niche" argument. Forget 4K, what % of people out there actually even rip Blurays compared to stream videos from the Internet?

If I get $1 for everytime I see that on PC tech forums I will be rich by now.

So you believe we've reached the end of technology then or what?

 

[StrangerGuy]

So you believe we've reached the end of technology then or what?

Surprise, tech has diminishing returns in real world utility?

More news at 11.

 

[rchunter]

Yet another tired "niche exist so it isn't niche" argument. Forget 4K, what % of people out there actually even rip Blurays compared to stream videos from the Internet?

If I get $1 for everytime I see that on PC tech forums I will be rich by now.

We will see... you might be right about the majority not needing it. I'd say within 5 years most enthusiasts on this forum will be running it. Especially those that need to move around large amounts of data, streaming 4k to their home theaters, etc.

 

[ksec]

Most people don't need it, that is true. Because the future is wireless, 802.11ax and 802.11ad v2 ( 60Ghz ) will be most you need. The first will provide first true 1Gbps wireless speed at home, the second will give you 10Gbps speed in direct Line of sight connection. When i say most that is likely to be 90% of users, who will be using Mobile Phones, Tablet, Laptop, and may be some desktop in 2018+.

Where 10Gbps will live are on Router, Switches and NAS. The problem is, when I say i don't see how it gets any cheaper, is Hardware Manufacture will simply ignore the bottleneck and continue to use 1Gbps port on those devices. Especially on the consumer market since 1Gbps are a lot cheaper, as Consumer market are a lot more price sensitive. It is the same thing happening now where there are cheap 802.11ac router running with 100Mbps Ethernet LAN / WAN..........

Those who care about the speed bottleneck for Router and NAS are geeks / enthusiast or Business / Enterprise. Since Business and Enterprise will not buy equipment from Consumer market ( or rarely ) their needs may never be filter down to us. Or at least not until very late in the cycle.

 

[Mir96TA]

I think wired Ethernet technology is dying breed in consumer market.
I can see more devices are WiFi based, and this trend is growing trend day by day.

 

[Eug]

Lately, some of the products I have bought have been aimed at the consumer to small and mid-sized business market, not pure consumer products. I'm fine paying these mid-range prices, but obviously I don't need to buy enterprise level equipment. For example:

All-in-one printer/scanner/copier/fax: Samsung ProXpress M3870
NAS: Synology DS413
Main network switch: Zonet ZFS3324e 24-port rackmount Gigabit switch

My Synology DS413 NAS only has one Gigabit connection, but it was from 3 years ago. It should be noted that the 2016 version, the DS416, includes 2 Gigabit connections with link aggregation, because Gigabit is too limiting. This model is classified as a "Value" model, not a high end model. It would be so much easier to just have a single 10GigE (or NBaseT) connection, with corresponding switches and client devices.

tl;dr

I sometimes buy small-to-medium-sized business peripherals for my home how, since low end consumer products are too limiting. I'd be willing to pay these mid-range prices for 10 Gigabit networking equipment.

 

[rchunter]

I'm pretty sure synology has already released at least one model with 10GbE. EDIT - Yeah, here it is https://www.synology.com/en-us/products/DS2015xs

 

[Eug]

I'm pretty sure synology has already released at least one model with 10GbE. EDIT - Yeah, here it is https://www.synology.com/en-us/products/DS2015xs

Yes I know, but I don't need an 8-bay $1500 NAS for my house. Those units are aimed at mid to large sized businesses. And plus, that still doesn't help in terms of affordable switches and 10GigE clients.

What I'm saying is I want to see stuff like $500 small business NASes get 10GigE by 2020, along with sub $100 switches, and personal computers with 10GigE built-in.

 

[sdifox]

Yes I know, but I don't need an 8-bay $1500 NAS for my house. Those units are aimed at mid to large sized businesses. And plus, that still doesn't help in terms of affordable switches and 10GigE clients.

What I'm saying is I want to see stuff like $500 small business NASes get 10GigE by 2020, along with sub $100 switches, and personal computers with 10GigE built-in.

8 bay nas is definitely not aimed at mid or large corp. Maybe soho.

 

[rchunter]

Yes I know, but I don't need an 8-bay $1500 NAS for my house. Those units are aimed at mid to large sized businesses. And plus, that still doesn't help in terms of affordable switches and 10GigE clients.

What I'm saying is I want to see stuff like $500 small business NASes get 10GigE by 2020, along with sub $100 switches, and personal computers with 10GigE built-in.

Easily by 2020. Probably long before that..

 

[Eug]

Easily by 2020. Probably long before that..

Hope so.

8 bay nas is definitely not aimed at mid or large corp. Maybe soho.

I hear where you're coming from, but I don't think I've ever seen a SOHO setup with a NAS like this one. It's one that's expandable to 20 bays, for example.

In fact, in AnandTech's review, they gave it an "enterprise" tag (even if that might be pushing it).