Forecast for 10GbE

[Matioupi]

Hello,

 

when do you think 10 Gb/s will overcome/replace the current 1 Gb/s lan ports on virtually all motherboards/systems that are sold at date ?

Do you have any doubts that it will ever happen ?

Do you have any doubts that it will be available through 10GB-BaseT  "RJ45" standard ?

 

[littlepigboy5]

2 minutes ago, Matioupi said:

Hello,

 

when do you think 10 Gb/s will overcome/replace the current 1 Gb/s lan ports on virtually all motherboards/systems that are sold at date ?

Do you have any doubts that it will ever happen ?

Do you have any doubts that it will be available through 10GB-BaseT  "RJ45" standard ?

 

when average consumer internet speeds exceed 1gbit which is a long way away, because on average it's about 15 mbps

[Orangeator]

1 hour ago, littlepigboy5 said:

when average consumer internet speeds exceed 1gbit which is a long way away, because on average it's about 15 mbps

This, unfortunately the max speeds I can get are 15mbps unless I pay $120 a month for internet with a low bandwidth cap.  This is what happens when you live just outside a really large city.

[desertcomputer]

When our download link pass 100mpbs first mate ... most people don't even have 20mbps. 

[Walker998]

not anytime soon

 

for one while the uk has unlimited data we have much lower bandwidth then a large amount of americans (i have 25mps)

 

seriously, by 2020 everyone will have at least 10mps here

 

EDIT: people have much lower bandwidth than i expected before someone corrects me

[Bananasplit_00]

when everyone has access to Gigabit speeds, so maby 5-10 years?

[Matioupi]

thanks for the interest in answering.

 

So far, you all seems to think that the internet speed is the key limiting factor.

I do have about 20 Mb/s internet, but still 10Gb/s LAN is really great for archiving, and professional storage optimizations through high speed NAS/SAN)

 

If the mean for WAN connections is 15Mb/s right now (didn't knew that), then why 1Gb/s would be the standard...

[DXMember]

Skylake EP is supposed to have that

[Walker998]

considering gigabit internet has been around since the early 2000's it will be interesting to see a graph on progress of higher speeds being more and more available 

[Razor512]

1 GbE has been obsolete for the past 10+ years, and each year, it has simply become exponentially more obsolete. 10GbE should have been the home network standard many years ago, router makers are unwilling to spend the 30 cents or so on a 10GbE switch, From a BOM standpoint, a 10GbE switch does not take much more in production.

[Linearburn]

I think you will start to see 10g creep in to the consumer realm with in the next 2 years there is currently a desktop board with built in 10g however 3-5 years before its common place unless something cheeper and faster comes along first

[Linearburn]

1 hour ago, Razor512 said:

1 GbE has been obsolete for the past 10+ years, and each year, it has simply become exponentially more obsolete. 10GbE should have been the home network standard many years ago, router makers are unwilling to spend the 30 cents or so on a 10GbE switch, From a BOM standpoint, a 10GbE switch does not take much more in production.

well kinda the processing power 10g needs in multiport is more honstly a rasberry pi 3 has enough scons to do it so I would expect 300 buck for the first consumer 8 ports.

[Matioupi]

6 hours ago, Linearburn said:

well kinda the processing power 10g needs in multiport is more honstly a rasberry pi 3 has enough scons to do it so I would expect 300 buck for the first consumer 8 ports.

according to a graph found here :

 

http://www.anandtech.com/show/8718/the-samsung-galaxy-note-4-exynos-review/3

 

describing memory bandwidth of the ARM Cortex A53 that is empowering the rpi3, it can be seen that memory bandwidth is somewhere between (depending on size of transfers) 16 and 40 Gb/s so it will probably be a little short for an 8 port switch.

 

By the way, for 10 GbE users, this graph questions me about what is seen on this guide (and following parts)  :

 

http://www.cinevate.com/blog/confessions-of-a-10-gbe-network-newbie-part-1-basics/

 

stating that buffers should be tuned at maximum values for improved throughput;

 

I did actually observed the opposite (with intel x710-da4 and x520-da2 cards) where reducing the buffer a little bit from default values would improve the throughput, but maybe the pretty old 10GbE switch in the middle is at play here with the same effect than on the above (and I don't have time testing without the switch)

[leadeater]

11 hours ago, Matioupi said:

Hello,

 

when do you think 10 Gb/s will overcome/replace the current 1 Gb/s lan ports on virtually all motherboards/systems that are sold at date ?

Do you have any doubts that it will ever happen ?

Do you have any doubts that it will be available through 10GB-BaseT  "RJ45" standard ?

 

10GBase-T already exists and I use it at home, Intel X540-T1.

 

It will happen, data bandwidth has to keep up with data storage size increase. Just look at USB speeds ever since 1.0, or IDE/SATA/SCSI/SAS.

 

It will happen when Intel include 10Gb in the standard desktop chipset and unlikely any sooner.

[skywake]

15 hours ago, littlepigboy5 said:

when average consumer internet speeds exceed 1gbit which is a long way away, because on average it's about 15 mbps

This doesn't have much to do with it. Internal network speeds and internet connections are not one and the same. Gigabit became a pretty standard feature on motherboards by around 2005. They were a feature on premium boards for a few years before then. It's not like 100Mbps+ internet was commonplace back then.

 

The way I see it networking on motherboards is the same deal as it is with any other I/O. A new thing demanding those speeds can push it along but once it becomes cheap enough they'll include it whether there's a demand or not. Like when SATA went from 1.5Gbps to 6Gbps in 5 years before SSDs were even on the scene despite the average HDD running at ~800Mbps upto ~1.4Gbps or so. There was no demand for it at the time but they could deliver it so why not? Same deal with NICs.

 

Currently a 10Gbps NIC is a couple of hundred bucks. When it drops down to ~$50 or so? Then they'll start including it regardless of whether you "need" it or not

[Razor512]

The issue is that there is no need for the 10 GbE cards to be so expensive, they don't cost that much to make. they are overpriced simply because they are seen more as an enterprise product, where everything has profit margins that will put apple to shame. for example

http://electronics360.globalspec.com/article/3509/juniper-networks-acx2000-universal-access-router-teardown

 

for 10GbE to enter the home setting as the default standard replacing 1GbE entirely, a networking company simply needs to break ranks from the price fixing, and significantly cut the price. This will cause them to see a massive boost in sales people scramble to upgrade their home networks because it is now at an affordable level where 8+ wired devices can be on 10GbE for a similar price to 1GbE.

 

With all of the other cheap, high throughput technologies, there is really no excuse for the steep pricing.

[Stefan1024]

10GBit/s over ethernet just needs a lot of electrical power. Even the 1GBit/s NIC consume quite a bit. So unless you really utilize the bandwith it's a bad idea to just use 10GBit/s as default. But I'd love to have a >1GBit/s link to my NAS.

[skywake]

25 minutes ago, Razor512 said:

The issue is that there is no need for the 10 GbE cards to be so expensive, they don't cost that much to make. they are overpriced simply because they are seen more as an enterprise product, where everything has profit margins that will put apple to shame

I think you've got it backwards. They're an enterprise product only because it's too expensive to sell to the average consumer. Especially one who probably doesn't have any immediate need for it. The price is going down it's just not going down as fast as it did for Gigabit. Mostly because there's less of a need for it than there was for Gigabit. Less need means less demand which means the cost doesn't go down and they don't start including it on everything.

[leadeater]

Also don't forget manufacturing cost is not the only factor in pricing, in fact it can often only be a small part of it. Product research and development is extremely expensive and return on investment needs to me made, this is why early sales of new technology is very expensive.

 

The only reason we have 10Gb, 40Gb, 100Gb ethernet standards is from costly research that all the networking companies contribute to, directly or indirectly. Once the technology is standardized enough they then also have to develop product lines that include this new technology.

 

The cost of actually making a product is often extremely small compared to the above costs.

[Razor512]

The issue is that 10GbE has been out of rover a decade now. There are certain products that kinda get a pass on the astronomical profit margins, for example CPUs and GPUs. this is because they can cost a lot in R&D for a product that will only be viable for a short while before it becomes outdated.

 

10 GbE has had ample time too come down in cost.

From a raw BOM standpoint, it cost more to make a raspberry pi than it does to make a 10GbE NIC

 

From an engineering standpoint, it is far more complex to implement a USB 3 controller, and astronomically more complex to design a fully functioning motherboard that will offer PCI-E 3.0 and various onboard devices, yet we can buy modern current gen socket motherboard for less than $50 if we wanted to. (New chipsets come out more often than new Ethernet standards, they are more complex than a basic 10GbE controller, and require more work to implement due to the need for length matching, controlled impedance, heavy via stitching, many PCB layers, and so on.

 

Compared to everything else, if real market forces were at work, we would 10GbE NICs  at electronic stores for $10-15, you would likely see them in bargain bin $100 2TB NAS devices simply because they are so cheap to make it that companies will take a "Why not" approach; it will essentially become a check box feature.

 

 

With computing, hardware has typically lead software development, for example, would anyone have attempted to create netflix, and youtube if we were still using 2400 baud modems to get online?

[Matioupi]

aside from all the BOM / price discussion which is interesting, I think there are other more practical" key factors

 

• Power consumption has been mentioned and plays its role, I've digged that a little mode : when used at gigabit only, a X550-T2 is given for 5.5W average 6.4W max (turns to 11.2 / 13W at 10GbE)

(http://www.intel.com/content/www/us/en/ethernet-products/converged-network-adapters/ethernet-x550-brief.html)

while a dual gigabit port http://ark.intel.com/products/84804/Intel-Ethernet-Server-Adapter-I350-T2V2 is given for  4.4W max

If you scale "per Gb/s" 10Gb NICs are more efficient, but they are still worse at 1Gb/s by about 45%  (of a small value) and at the end of the day, the bill will be bigger so you have to sell more GB (TB) to earn more money (which is the trend...)

• The physical layer also plays it's role. Are there any statistics/polls (could be a separate thread) about he actual percentage of users that are doing WiFi only vs. thoses that are using wired connections (at home / at work). I don't think 10GbE WiFi is quite there yet... and therefore generalized WiFi usage may slow down quick adoption of faster network standards at a large scale in a similar way that it did for 10/100 to 1Gb transition. A lot of consummer content/service providers will stick the bandwidth requirements of their products (games, video, ...) to what can be handled by WiFi. Only pro or wired (which I guess is a small percentage) will benefit turning 10GbE.

• For the good points, RJ45 10GB-base T was far from being an easy way to go for 10GbE up to now. Almost all pro switches and NICs were SFP+ or CX4 only on the physical layer. Now, we have the intel X540 and X550 series for RJ45 based NICs, plus some "comsummer/small business" 10bE base T / mixed physical player switches, and this will allow people to turn more smoothly to 10GbE. Even if cabling was only cat 5e, tests tends to show that 10GbE is quite well supported for reasonnable distances.

[Tosa]

On 14.3.2016 at 8:50 PM, MrUnknownEMC said:

When our download link pass 100mpbs first mate ... most people don't even have 20mbps. 

I have 100 mbps, I can get 500 if I pay up, and I don't believe my ISP offers anything below 25 mbps. I've heard about someone on the news a good while back that's participating on some kind of test to see how well they can make 1 gbps work through copper cables.

[desertcomputer]

4 hours ago, Tosa said:

I have 100 mbps, I can get 500 if I pay up, and I don't believe my ISP offers anything below 25 mbps. I've heard about someone on the news a good while back that's participating on some kind of test to see how well they can make 1 gbps work through copper cables.

Well you could get 10gpbs internet if you pay thousands of dollars for the setup. For most consumers the fastest speed is 100mbps which only for people who has nbn which is less than ~10% (Not sure about actual statistics).