If my sums are right then that's about 43690.67 times faster than my first modem. Blimey. Is my life 43690.67 times better for this technology though? Probably not.
Psst. We've got 400Gb/s Ethernet working - but don't tell anyone
Californian company Ixia has shown what it claims is the the world’s first functioning 400Gb/s Higher Speed Ethernet test rig, based on the IEEE P802.3bs protocol. The company says that this bandwidth is sufficient for 50,000 simultaneous high definition Netflix video streams. The 400GbE Jumpstart Test system provided wire …
-
-
-
-
Thursday 26th June 2014 01:42 GMT wbaw
Re: First modem
You had a modem? Luxury. Here in Yorkshire ...
I use to have to copy games on to audio tape & swap them in the playground. The school's (single) computer was a BBC B & nobody else could afford those, so it was only educational software on it.
And you try and tell the young people of today that ..... they won't believe you.
-
Thursday 12th June 2014 12:41 GMT Anonymous Coward
@Dr Who
I think you got your maths wrong, unless your first modem was an ADSL ....
400Gb/s is 400,000,000,000 bits per second so if you modem was as slow as mine at 1200bps that would be over 30 million times faster. 56k modem would be over 7 million times faster.
That's two orders of magnitude more, makes life even better ... :)
-
Thursday 12th June 2014 14:51 GMT Moonshine
"Is my life 43690.67 times better for this technology though?"
I asked myself a similar question when I purchased a Raspberry Pi last year and calculated that BBC Basic runs about 2,500 times faster on the Pi (running RiscOS) that on the original beeb (and the Pi 1/20 the price in real terms).
Is the Pi 2,500 better than the BBC Micro? Not quite ;)
-
-
Thursday 12th June 2014 22:11 GMT Nick Ryan
Re: Entire article fails to mention the other factor...
Distance
Please restate your question using proper and correct measurements. It makes a lot more sense in Linguine.
For your reference and correctional education: http://www.theregister.co.uk/2007/08/24/vulture_central_standards/ or for the slide rule shy (*): http://www.theregister.co.uk/Design/page/reg-standards-converter.html
* I'll be buggered if I know how to use a slide rule either.
-
Friday 13th June 2014 14:31 GMT Michael Dunn
Re: Entire article fails to mention the other factor... @Nick Ryan
" I'll be buggered if I know how to use a slide rule either." Poor chap! I got my first one at age 11 - celluloid glued to paper glued to wood. Necessary for School Certificate Physics and Chemistry.
Plastic eventually took over, and I've still got a Faber Castell I used much later.: folded dual scale (so 10" was effectvely 20") with reciprocal scale in the centre of the slide, sin, cos, tan scales on the back (not to mention the bibulous shin, cosh, and tanh), and a little KW to HP coversion offset on he cursor..
Who needs a calclator?
Yes, I know, some Japanese whizz-kid with a soroban (abacus) can do it faster!
-
-
Friday 13th June 2014 07:27 GMT Tom Samplonius
Re: Entire article fails to mention the other factor...
It is fiber optic, so several kilometers at minimum, depending on the output of the transceiver.
Distance is not really a problem, as there are people doing 400Gbps on fiber today for 100+ kilometers, but as 40 discreet 10Gbps channels. This is just standard WDM in use for 10+ years. The only thing special about 400Gbps ethernet, is that the WDM is done at the media layer, so it appears as one logical circuit. Eventually, telcos will be able to chuck all of that WDM gear, and just use ethernet directly on the glass, right off their core routers.
-
Friday 13th June 2014 20:18 GMT Terry Barnes
Re: Entire article fails to mention the other factor...
"It's kinda an important detail. Right?"
Yes, but in a different context. This kit is capable of generating a datastream at a very high rate but it's not a transmission device. The transceiver of this device would be connected to some kind of transmission kit to leave the building. This bitrate is easily handled today by DWDM kit - mind-boggling data rates over fibre is unexceptional in telco core kit. In terms of distance - how long is a piece of fibre?
If what you're asking about this as an access line technology - out to an end user site - no. You could put the same telco DWDM kit in someone's house but they wouldn't like the bill.
-
-
Friday 13th June 2014 06:44 GMT TheSkunkMonk
I went from downloading images in 10-15mins to downloading high def video in less than one, sadly this has done little for making the world a better place. It's the same old story with large IT companies though always holding back hardware instead of producing the best they can, and the practices they have on the stuff they do release in the name of profit is absurd as well. http://hackaday.com/2013/03/18/hack-removes-firmware-crippling-from-nvidia-graphics-card/
-
-
-
-
Thursday 12th June 2014 11:15 GMT Jellied Eel
Re: Reg's standard for this?
Bindun-
"The company says that this bandwidth is sufficient for 50,000 simultaneous high definition Netflix video streams."
So 400Gbps = 50KNvs. Which may not be a lot depending on how many of Netflix's 44m subscribers stream concurrently. I suspect industry may skip 400Gbps as they pretty much did 40Gbps in favor of 1Tbps Ethernet.
-
Thursday 12th June 2014 15:30 GMT Lusty
Re: Reg's standard for this?
"I suspect industry may skip 400Gbps as they pretty much did 40Gbps"
I think you'll find 40Gbps is incredibly popular among those who need it. The reason you may not have seen much of it is that very few people do need it. 10GbE is sufficient for the vast majority of infrastructures with 40GbE and 100GbE only really necessary when connecting up lots of large switches, for instance in data centre use or at very large companies. It's occasionally useful on very fast flash based SAN too, although this is also pretty rare.
-
Thursday 12th June 2014 19:43 GMT Jellied Eel
Re: Reg's standard for this?
"I think you'll find 40Gbps is incredibly popular among those who need it. The reason you may not have seen much of it is that very few people do need it."
I work on the supply side and haven't seen a request for 40G in a long time. That's on the WAN rather than LAN though. Reason I'm sceptical is historical. 40G was ok for older systems that worked on Nx10G wavelengths, newer systems use Nx25G which is one of the reasons why 100G is standardising around 100GBase-xR4. On the line-side flexible grids (ITU-T G.694) using PM-QPSK are running 500Gbps and 9Tbps+ on a single pair. So generally 40G isn't a convenient or efficient fit with modern muxes. 400G may work as in interim 4x100 solution but again less efficient it you end up wasting 100G on a line card. 1Tbps is possible now with 2xOSCs between compatible telcos, which generally means another DTN-X user. More info here-
https://www.infinera.com/technology/files/infinera-IEEE-Meeting-Superchannels.pdf
-
-
This post has been deleted by its author
-
-
-
-
-
Thursday 12th June 2014 11:17 GMT Geoff Campbell
Re: I want a high speed version of 10base2
The whole CSMA/CD system that 10base2 uses starts to break down quite badly at very low traffic speeds. This was why 4Mbps Token Ring would piss all over 10Mbps Ethernet back in the day, as soon as you got a few systems using the network in a spirited fashion.
GJC
-
Thursday 12th June 2014 12:45 GMT Anonymous Coward
Re: I want a high speed version of 10base2
"The whole CSMA/CD system that 10base2 uses starts to break down quite badly at very low traffic speeds"
Yeah I know , but a man can dream can't he? :o) Though I honestly don't see the point of 10BaseT - its essentially just a serial link with the overheads of a network protocol (hardware and software). Why not just use a proper serial link and be done with it.
-
Thursday 12th June 2014 13:33 GMT Anonymous Coward
Re: I want a high speed version of 10base2
well - there are 2 additional things Ethernet gives you vs a serial link:
1) share the same collision domain between multiple devices. Not really an issue now all ports are switched - and once you go full duplex collisions go away altogether.
2) share the same broadcast domain/IP subnet between multiple devices. More important - anything that depends on mDNS (e.g. Apple Bonjour) or UPnP (e.g. DLNA) will break once you put a subnet boundary in the way.
-
Thursday 12th June 2014 14:28 GMT Anonymous Coward
Re: I want a high speed version of 10base2
"well - there are 2 additional things Ethernet gives you vs a serial link:"
Sure, but that can be done at the router end - you don't need it at the client. The client shouldn't need to bother with MAC addresses et al, just know its IP address. I don't see any technical reason why client PCs couldn't simply be connected to the bridge/router via a high speed serial link perhaps along the lines of USB or SATA or perhaps something entirely new. But without the overhead of ethernet.
-
-
-
-
-
Thursday 12th June 2014 20:27 GMT JeffyPoooh
Re: I want a high speed version of 10base2
AC: "...your PC probably doesn't have a built-in ONT."
My Bell Aliant FibreOP ONT is screwed to the plywood panel that I installed for that purpose in the basement. The nice technician bolted the ONT to the plywood and connected the fibre (FTTH). He also installed a battery backup system, since the same fiber provides POTS (e.g. 911 service). Also a wifi router was thrown in with the deal, but I've since installed two more to fill-up the entire 2.4 GHz band with hotspots. Even one on 5.24 GHz. There a growing network of Cat 6 and a lovely Gb switch.
Finally, I'm in the top 2% of Canada with 175 / 30 Mbps and 3ms ping to the nearest Speedtest server.
-
-
-
Thursday 12th June 2014 11:48 GMT AndyFl
4k streaming on mobile networks?
From the article:
"With increasing backhaul pressures on mobile phone networks that will want to stream large numbers of 4K video streams to subscribers, the bandwidth for each level of infrastructure matters."
I don't think ANY cellular network operator wants you to stream 4k video streams to clients at the moment and probably not for many years. One 4k stream will easily take 15Mbit, get a couple of subscribers on a site and watch everyone else grind to a halt. In any case I think the 7GB or so per hour traffic would eat through a data plan very quickly. I also suspect would break the terms and conditions of the '3' unlimited plan - assuming you can get that download speed on their network anyway.
The future for mobile video is in rate-adapting CODECs which adjust the stream throughput to the screen size and available backhaul capacity.
Andy
-
Thursday 12th June 2014 12:35 GMT Anonymous Coward
Re: 4k streaming on mobile networks?
"The future for mobile video is in rate-adapting CODECs which adjust the stream throughput to the screen size and available backhaul capacity."
Thing is, we're already a year into devices fully capable of both decoding and displaying 1080p content, and that's a hog in and of itself. Not only that, network tech heads are beginning to throw up the white flag, saying we're approaching the physical limits of wireless data technology.
Add to that the fact that content providers insist on live streaming without any possibility of a download (since ANY download, no matter how encrypted, can be broken) and you're approaching an impasse.
Something's gotta give soon. Either content providers give up on trying to stream HD video routinely or they give up on trying to block downloading.
-
-
-
Thursday 12th June 2014 12:07 GMT Mage
Mobile Fantasy
"mobile phone networks that will want to stream large numbers of 4K video streams to subscribers"
There is no mobile tech than can stream more than approximately ONE 4K stream on a Mast sector, and that needs largest size channel and perfect signal. Typically you only need 400Mbps (peak) for a multi-sector mast with 20MHz channels. Most masts will struggle to use up 100Mbps of backhaul peak. The Headline Mobile speeds are frequently on very unlikely configurations and 1 user. Probably 1Gbps backhaul "future proofs" a mast.
HFC (DOCSIS 3.0 cable cabinet fibre fed), VDSL, FTTx etc benefit from this. Mobile not at all.
-
Thursday 12th June 2014 13:36 GMT Anonymous Coward
Re: Mobile Fantasy
Yes - the cost differential between fixed and (macro-cellular) wireless access per gigabyte of data shifted is immense. A while ago it was about 1c per Gig on fixed and $1 per Gig on wireless. Interestingly at the same time FLASH memory was about $1 per Gig.
frankly the wireline providers *need* video services to defend themselves from the mobile guys. The statistical gain on web browsing etc. is just too good - if that was all we used the Internet for then the wired network would wither on the vine.
-
-
-
Friday 13th June 2014 20:37 GMT Terry Barnes
Re: Way ahead of the access layer
Data traffic is more than just the sum of consumer broadband connections. Thing about the amount of data that will be flying around as the result of Big Data and Internet of Things. IBM are delightedly telling people at the moment that 90% of all the data ever created was created in the last two years - and that data wants to go places.
-
-
Friday 13th June 2014 20:40 GMT Terry Barnes
Re: 400GB
" 400GB
Wow. That is crazy stupid fast"
It would be - but this is 400Gb. Units are important.
"Too bad my avg home speed is still between 1-7MB (depending on the moon cycles and coin flips and horse races)."
You could have 400Gbps to your home today via DWDM. You might not be too happy about the bill though.
-