10 Gigabit networks

Doesn't seem right but then whats stopping one of those chaper manufactures from doing it, like you can get a 5 port 1 gigabit switch from tp-link for £14, pcie cards are £7, i would be willing to spend three times more than that for something at 10 gigabit.
 
Radiation said:
Doesn't seem right but then whats stopping one of those chaper manufactures from doing it, like you can get a 5 port 1 gigabit switch from tp-link for £14, pcie cards are £7, i would be willing to spend three times more than that for something at 10 gigabit.
Click to expand...

10Gbit/Second is a hell of a lot more data to shift around than 1Gbit/Second.

The extra processing involved in the time is a pretty cool feat.
 
Radiation said:
Doesn't seem right but then whats stopping one of those chaper manufactures from doing it, like you can get a 5 port 1 gigabit switch from tp-link for £14, pcie cards are £7, i would be willing to spend three times more than that for something at 10 gigabit.
Click to expand...

Q) How long after gigabit ethernet was introduced as a technology did the likes of TPLink bring out their garbage switches?

A) Almost a decade.
 
Radiation said:
Doesn't seem right but then whats stopping one of those chaper manufactures from doing it, like you can get a 5 port 1 gigabit switch from tp-link for £14, pcie cards are £7, i would be willing to spend three times more than that for something at 10 gigabit.
Click to expand...

There isn't enough consumer demand for 10Gb/s speeds at 3-4 times Gigabit prices for the companies that have the patents/licences for 10Gb to make up for the huge cut in profit margins they'd take on kit sold to enterprise customers.

Or in other words, there's more money to be made skimming off the top than racing to the bottom.
 
Last edited:
The thing is though, regardless of whether it's needed or not if it was priced the same as 1 gigabit it would replace them pretty quickly as there would be little point in keeping the old technology, it's better to sell more for less than less for more as it will benefit more people in the long run.
 
paradigm said:
Q) How long after gigabit ethernet was introduced as a technology did the likes of TPLink bring out their garbage switches?

A) Almost a decade.
Click to expand...
I think my Gigabit switch is TPLink...it may not saturate 1 Gbps all the time but I wouldn't call it garbage. Cheap & cheerful, certainly - couldn't really go wrong for £15.
 
Radiation said:
The thing is though, regardless of whether it's needed or not if it was priced the same as 1 gigabit it would replace them pretty quickly as there would be little point in keeping the old technology, it's better to sell more for less than less for more as it will benefit more people in the long run.
Click to expand...

The amounts that will have gone into R&D for enterprise grade kit will be astronomical. There is no consumer grade 10GbE kit yet, as such you won't be getting consumer pricing anywhere near the "near future".

I think you greatly misunderestimate the actual markets for enterprise kit.

DragonQ said:
I think my Gigabit switch is TPLink...it may not saturate 1 Gbps all the time but I wouldn't call it garbage. Cheap & cheerful, certainly - couldn't really go wrong for £15.
Click to expand...

Compared to proper switch fabric, intelligent switching fabric that can sustain the throughput to all ports on a hundred or so ports in a switch stack simultaneously, it's utter cheap garbage. And this is the entire point, the manufacturers are charging astronomical amounts for their 10GbE switching because the units themselves are incredibly complex, the logic within them is complex, and the research into them was complex. A TPLink switch (even if a 10GbE one were to exist) would be so far off the scale, it's incomprehensible to even compare the two.

The equivalent of purchasing a little tykes plastic bubble car rather than a rolls royce.
 
Last edited:
paradigm said:
Compared to proper switch fabric, intelligent switching fabric that can sustain the throughput to all ports on a hundred or so ports in a switch stack simultaneously, it's utter cheap garbage. And this is the entire point, the manufacturers are charging astronomical amounts for their 10GbE switching because the units themselves are incredibly complex, the logic within them is complex, and the research into them was complex. A TPLink switch (even if a 10GbE one were to exist) would be so far off the scale, it's incomprehensible to even compare the two.

The equivalent of purchasing a little tykes plastic bubble car rather than a rolls royce.
Click to expand...

None of which is required for the home user - I don't need 100 ports in a stack, TBH I doen't even need a switch at the moment - I am only looking at connecting my main PC to my WHS box - I regularly transfer large (>15Gb) files and prior to going away I will often push around 1Tb of data to my 'away' box. I don't need an industrial setup - just something faster than the current Gigabit environment that is seriously starting to limit what I do :p

All that is really needed is for the big guys to start pushing 10Gb ports on their mobos and suddenly no-one would bat an eyelid - just look at how quickly Sata3 went from "you'll never use that - nothing uses all of Sata2 yet..." to "Oooooo - I guess we need something faster already!"

Is there anything preventing direct connection between the 2 machines without a switch?
 
There is nothing preventing just popping a crossover cable between two PCs.

Why should the companies, who have developed 10Gbit, waste money putting what they can fetch large sums of money for into home kit?
 
paradigm said:
Compared to proper switch fabric, intelligent switching fabric that can sustain the throughput to all ports on a hundred or so ports in a switch stack simultaneously, it's utter cheap garbage. And this is the entire point, the manufacturers are charging astronomical amounts for their 10GbE switching because the units themselves are incredibly complex, the logic within them is complex, and the research into them was complex. A TPLink switch (even if a 10GbE one were to exist) would be so far off the scale, it's incomprehensible to even compare the two.

The equivalent of purchasing a little tykes plastic bubble car rather than a rolls royce.
Click to expand...
Yeah but you're talking about enterprise-grade managed switches, right? Not really the same as simple, unmanaged switches for home networks. Like you say, consumer grade 10 Gbps equipment doesn't even exist.
 
So why argue that 10GbE is too expensive? Just understand that the tech is firmly in the hands of enterprise only for the forseeable future.
 
cavemanoc said:
All that is really needed is for the big guys to start pushing 10Gb ports on their mobos and suddenly no-one would bat an eyelid - just look at how quickly Sata3 went from "you'll never use that - nothing uses all of Sata2 yet..." to "Oooooo - I guess we need something faster already!"
Click to expand...

An Intel single port 10Gb server NIC is going to set you back £400 - that's quite a drop needed for them to appear on bog standard motherboards.
 
Not really - as mentioned, a combination of competitive licensing and mass production soon bring the price down, I'm willing to bet that we see the first onboard 10Gb over the next 2 years - I know I will be picking up a couple of cards over the next year and setting up a point to point between my server and the main PC - when I find cards at the £150 mark I'll be clicking buy!

You only have to look at the cost of the server boards/chips to see the mark ups involved - obviously the dual/quad socket boards have different features to the standard home kit, but that doesn't mean that cheap home versions of them aren't available (the features - not the sockets ovb.)

I am often quite surprised how far behind the tech gets on occasion - I remember picking up my cable router a few years ago and being amazed by how few of them offered gigabit ethernet.
 
Last edited:
paradigm said:
So why argue that 10GbE is too expensive? Just understand that the tech is firmly in the hands of enterprise only for the forseeable future.
Click to expand...


And only certain applications within the enterprise at that. I agree there is no demand for it in the mainstream. It'll be a long time yet before we see it at the prices 1gbit was even 2-3 years ago.
 
I'd gamble on longer than 3 years before its anything like comparable to 1GbE in the home. Until we start seeing 10GbE onboard NICs in servers there is no chance in the consumer market.

Wait for 40GbE to be the new enterprise 'toy' then costs associated with 10GbE will drop much faster.
 
You must log in or register to reply here.
Back
Top Bottom