New Data Transmission Speed Record

An anonymous reader writes "Gizmag is reporting that a team of German and Japanese scientists have collaborated to shatter the world record for data transmission speed. From the article: "By transmitting a data signal at 2.56 terabits per second over a 160-kilometer link (equivalent to 2,560,000,000,000 bits per second or the contents of 60 DVDs) the researchers bettered the old record of 1.28 terabits per second held by a Japanese group. By comparison, the fastest high-speed links currently carry data at a maximum 40 Gbit/s, or around 50 times slower."

2.56 Terabits = ?

Anti-Slashdoting for a webhost.

Terabits per second!?

What kind of measurement is that? Why can't we use something everyone understands like u-haul trucks full of dvd's driving 100 kilometers per fortnight*10^(-6)?
For the uninitiated, that's a microfortnight.

Re:Terabits per second!?

A measurement everyone understands? Like how much porn?

Re:Terabits per second!?

A measurement everyone understands? Like how much porn?

Well, they did say 60 DVDs. So depending on your proclivities, that's ... what ...

18 hours of phone sex ads + 720 'standard' sex scenes

180,000 pop shots on compilation

240 hours of midget porn

6,000 discrete pee sequences

No less than 100 gang-bang sequences

500 bukkakke events

600 different MILFs

1 million or so jpeg images

I mean, these are known measures. ;-)

How much in terms of

How much it is in terms of "a station wagon full of magtape" or Library of Congress.

http://en.wikipedia.org/wiki/List_of_strange_units _of_measurement#Books_and_Bible [wikipedia.org]
http://en.wikipedia.org/wiki/List_of_strange_units _of_measurement#Library_of_Congress [wikipedia.org]

So, that's OC-What?

Someone have the conversion factor handy?

-jcr

60 DVDs per second

(equivalent to 2,560,000,000,000 bits per second or the contents of 60 DVDs)

Wow, converting to MPAA units that's 300 years of jail time per second! Smokin!

Re:60 DVDs per second

No, no, no... It's $37,500,000 per second. ($125,000 per infringement)

Remember that the MPAA thinks in dollars.

Why is bandwidth measured in Kb

and storage in KB? It's not like we measure radio waves in cycles per second and sound waves in cycles per 8 seconds. What's the advantage?

Re:Why is bandwidth measured in Kb

One reason is because it's serial data generally, and you don't know exactly how many of those bits are going to be data.. (you could have start or stop bits, etc)... but I don't know the details of that, so I'll just mention my other possible reason.

It's that throughput is generally what actually matters when sending data. In other words, that how much actual payload is being send, minus any overhead. If you've got a decent amount of overhead, your actual throughput might be a bit less. So it makes more sense to talk about bandwidth in bits per second, so as not to confuse it with actual throughput.

Re:Why is bandwidth measured in Kb

You think wrong. Some quotes from the Wikipedia [wikipedia.org] entry:

In recent years, the use of a byte to mean 8 bits is nearly ubiquitous

Meaning even today it's not universal.

A contiguous sequence of binary bits in a serial data stream, such as in modem or satellite communications, or from a disk-drive head, which is the smallest meaningful unit of data. These bytes might include start bits, stop bits, or parity bits, and thus could vary from 7 to 12 bits to contain a single 7-bit ASCII code.

Here I think is the most revealing definition for the discussion in the present context.

The eight-bit byte is often called an octet in formal contexts such as industry standards, as well as in networking and telecommunication, in order to avoid any confusion about the number of bits involved.

Another site [uiowa.edu] says that:

* Pre-1965, and including the IBM 701, bytes were almost always 6 bits, though they weren't called that much then, but rather characters.

* 9 bits were sometimes used

* The PDP-6, PDP-10, and DECsystem 20 all supported changing the byte size with instructions from 1 to 36 bits (probably only some of those)

The latter reference, looking up the PDP-10 on Wikipedia, gives this quote:

Some aspects of the instruction set are still considered unsurpassed, most notably the "byte" instructions, which operated on arbitrary sized bit-fields (at that time a byte was not necessarily eight bits)

Actually it's 64

According to the numbers given, the old record is exactly 64 times slower. Hmmm. About 50. Right. Even I'm not too lazy to use a calculator. Personally, I would have given a metric like DVDs per second (which in this case is a better approximation of 544).

shattered? ok but...

ok, it shattered that record. but the male ejaculation has a genetic bandwidth of 8.4 x 10^6 GiB/s.

it'll be a while before they splatter that record.

(oh, and the scrambled word I have to type in to post this as an AC is "saturate"! yes indeed)

And so it begins

Well folks, time to gear up.

We know what happens when the Germans and the Japanese collaborate ;-)

 

Re:And so it begins

Well folks, time to gear up.

We know what happens when the Germans and the Japanese collaborate ;-)

Yeah, breaded and deep fried sushi, dark beer in small cups, and Teryaki sausages, LOTS of cabbage, and bubble-tea streusel.

Oh, the humanity of it all.

Not filling 1gb pipe

I've done some checking of aggregate SAN bandwidth in the last few months, and for the most part companies are not even saturating their 2 gbit Fibre Channel links (for that matter, most could easily run with legacy 1 gbit gear). Even the inter-switch links are often crusing well below 2 gbit. Funny to note that the big push is to release 4 gbit gear, for what conceivable purpose I have no idea.

It just seems to me that the real issue is access time, not bandwidth, though kudos to this team for an essentially meaningless achievement.

Filesystem and Ultra320 SCSI are our chokepoint

We have a 30TB EMC CX-500 with Brocade 2Gb FC backbone. The bench moves blocks of a few hundred GB to a dozen servers or less. We have never come anywhere near %50 utilization on the FC.

The transfers run about 4-6hrs and I was looking for choke points to shorten the time. The data simply won't go to disk any faster on the U320 SCSI bus. We consistently measure 20MBps max to disk, which makes sense. U320 means 320Mbps/8 = 20MBps. So I get the same max numbers for local disk-to-disk that I get for SAN-to-disk, and the same results regardless of OS. If this rate could be maintained, six servers doing the transfer should just about saturate the backbone, but the overhead of file access and FS management mean the max is only maintained for a moment as a few particularly large files come across. With lots of smaller files being copied, the average rate goes down to 2MBps.

If these servers had to be optimized for SAN-to-Disk transfer rate, they would have to have multiple SCSI controllers and HBAs, paired up on seperate PCI busses, and the data would have to be optimized with fewer/larger files.

Of course, the 2.5TBps link is of interest to ISPs and regional carriers not server labs, but I thought I'd throw in what we've seen on the utilization of a 2Gbps FC link in a SAN setup.

everybody needs this at home

for when you get this at your place :

http://www.onzin.nl/internetdownload/ [onzin.nl]

Way cool...

I'm calling my cable company to see if I can pay $25,000/month to get the data rate. Needless to say this will increase the downloads of my extremely intelligent blog.

Great Things Happen

when the Germans and the Japanese unite...oh wait.... There's a french joke somewhere in here but I just can't pull it out at the moment... i give up. -- J.

A 160-kilometer link?

Equivalent to 160,000 metres and 160,000,000 millimetres!

Basic requirement soon enough.

Vista will require this much bandwidth as a minimum, to download security patches and have them rendered in erotic 3D by the compulsory 4Tb graphics card, when it is released in THE YEAR 3000.

Try 10,000 Kilometers

The researcher assumes the transmission capacity on the large transoceanic traffic links will need to increase to between 50 and 100 terabits per second in ten to 20 years. "This kind of capacity will only be feasible with the new high-performance systems."

160 kilometers? large transoceanic traffic link? When 10,000 kilometers of non-repeated distance can be achieved, I'll be impressed. Until then it's nothing but a bragging right.

Well thats nothing

I can send data at about 300,000m/s (its called turning the light on)

fiber optics

"...researchers have now managed to squeeze more data into a single pulse by packing four, instead of the previous two, binary data states in a light pulse using phase modulation."

Is this similar to DDR memory where they pack info into the upward swing of the wave, and on the downward swing, as well as the troughs and peaks of the waves?

60 DVD:s per second...

...and in related news, the spokesman for the MPAA is currently unable to comment due to suffering a heart attack.

What did they send?

"...transmitting a data signal at 2.56 terabits per second over a 160-kilometer link..."

Ok Amazing speed etc etc etc but what did they send?
Was it encrypted?
Was it DRMed???

In the name of...

porn.... the germans uploaded 30 dvd's of their famouse Scheiße porn, and the japanese uploaded 30 dvd's of their bukkake favorites

faster than ram

i could be wrong, but isn't that like faster than the communications between parts in a computer including RAM? is there a computer that can feed that much data that fast?

Re:faster than ram

The target use for something like this is Internet backbone traffic, so the question is whether Cisco will be able to deliver a router that will keep the line busy. Cisco's web site says "the innovative 12000 Terabit System scales to 5 Terabits (Tbps) per second ".

The last mile...

...is all I care about anymore. A zillion bits per second in the lab or even in a backbone link doesn't do me a bit of good when my cablemodem has me restricted to 128k of upstream bandwidth. I just put together a killer computer system with more cpu, memory, RAM, and disk than I am likely to need anytime soon all for less than a kilobuck but decent bandwidth is still simply not available.

Close

That's almost fast enough to support the average Slashdot's readers requirements for pron.

So that's about 3.28 times of NetBSD's record

From this link: http://proj.sunet.se/LSR3-s/ [sunet.se] (This was from an old /. story)

NetBSD's Internet2 land speed record was (back at Sep'04):

# Network Distance: 28,983 kilometers
# Data transferred: 1831.05 Gigabytes (1966080000000 bytes)
# Time: 3648.81 seconds

Which equals 124.935 petabit-meters/second (1,966,080,000,000 * 8 * 28,983,000 / 3648.81)

Record in this story equals:

2,560,000,000,000 * 160,000 = 409,600,000,000,000,000 bit-meters/second = 409.6 petabit-meters/second

So that's about 3.28 times of NetBSD's record. Considering the fact that NetBSD's record was about 1.5 years old and they were using off the shelf hardware (Dell 2650 and Dell 650, Intel Pro/10GbE, routers were Cisco highend) except the routers, this new record is not all that impressive IMHO.

Single-channel only

To be sure, I believe this is a single-wavelength transmission record. For WDM (multiwavelength), I believe Alcatel's 2002 record of 10 Tbps over 3 x 100 km still hasn't been topped (Frignac et al, OFC 2002 [ieee.org]).

2.56 terabits per second?

whats that in libraries of Congress per second... is that the same as shuttle launches per second, or I am I getting confused with burning libraries of congress per shuttle launch?

Ooooh Multiple choice!

This will...
A.) Make internet (connection fees) 50x cheaper.
2.) Finally still western telecoms complaining about the difficulties of offering fast service.
3.) Be considered too expensive to implement until someone who can't be bribed or bough enters the telecom game.
4.) Never see the light of day.
5.) Be implemented perfectly showing the telecommunications industy has a commitment to quality!

Hardware?

I RTFA but didn't see any mention of hardware specs, so what did they use? What kind of disk can process that amount of information per second? Any speculations?

I have a feeling that if you tried hooking up my laptop, its 4200 rpm drive would pack up and walk out of the door, on principle alone ;)

Gizmag?

(snorts coffee out of nose)

Cant beat a good Gizmag

Until the kit can handle it...

I'm not sure what the utility of this is right now. In theory I could send 5.12 terabits/sec down a cable by getting the laser to flash on and off twice as quickly. That doesn't mean I've encoded any data or that I've been able to process that data at the other end in a meaningful way.

It's great that the transmission hardware is up to silly speeds, but until you can take that incoming data and packet switch/route it properly, until there are servers that can process even a tenth of that data in a meaningful way, this isn't something that will affect anybody, anywhere, other than for the people involved whose paper will get published.

Still pretty slow...

In fact, I'll bet a U-haul truck loaded with DVDs would still be faster.

Gee, that's fast....

The article says nothing about how long they can *sustain* it for. I mean, great. They can transfer 282 GByte in one second. If they can only do it for one second before the system breaks down, then I really don't care. Call me when they make a system that can transfer that data rate indefinitely.

In other news.....

In other news, it was reported that top officials of the RIAA and the MPAA were planning aggressive litigation to halt this progress citing 'Won't someone please think of the rampant piracy this technology will foster'. 'We believe that consumers cannot be allowed access to this kind of bandwidth'.

Cable company conflict of interest

DSL stinks and Cable Co's aren't interested in providing more unless there's more competition.

This is public muni FTTH FIOS is important . More transmission speed and better value.

The cable co's don't want anything to compete against their channels. That would be stupid. They are holding everyone back.

The internet backbone also doesn't want more transmission speed. They are just phone companies that want to gouge people.

If we could bypass all of them with a national fiber network.........

Which did they do?

Did they mean 2.56 terrabits or 2,560,000,000,000 bits? Unless you are a hard disc seller or unclear on your units, the two are very different.

Let me say that I am one of those tiresome people who think that there are 1,024 bits in a kilobit, 1,024 kilobits in a megabit and so on. KiB and its relatives are just an invention dsigned to confuse people and they serve no real purpose.

2,560,000,000,000 bits is about 2.32 terrabits and 2.56 terrabits is 2,814,749,767,107 bits. The two are not the same.

It's a bit like saying my cars top speed is 140 MPH or 287 KPH because car manufacturers have decided to redefine the kilometre. You can't. 1 Mile = 1.609 Km so my car does 225KM/hr.

Fellow geeks and pedants, let us get this right!

End pointless rant...

simple math

lets see, the new record is 60 DVD/s "ordinary" lines come in at 50 times slower, so still well over 1 dvd/second 60/50=x x=ludicrus speed Hey bob, uhh we need to move all of the internet to my house... and back again, by tues.

What kind of DVD ? BluRay ?

2.5 TB, 60 DVD ?
Let's see...

# rpncalc
rpncalc version 1.35. Copyright (c) 1993-2004 David Frey et. al.
This is free software with ABSOLUTELY NO WARRANTY.
For details, type `warranty'.
Type `quit' to quit and `?' to get a summary.
2.5 1024 dup dup * * *
1: 2684354560
60 /
1: 4.474e+07
1024 dup * /
1: 42.67

Well, Unless DVD can contain more that 40 GB of data (BluRay ?), there is a possibility of overflow in submitter's calculations.
My guess is that it's more like the quivalent of 600 DVD per second which has been transmitted.

Re:Digg Loses

Precisely why I read slashdot and fark, but not digg.

Slashdot has the non-time sensitive, most interesting news - with insightful or interesting comments.

Fark has the time sensitive or humorous news, with clever or funny comments.

Digg is somewhere in the middle, with the immature comments or spam I can find in an AIM chat room if I need it.

Nothing ever makes it out of the lab

into the real world, right? Why are they developing nuclear pebble bed reactors in laboratories, when I can't buy one at the 7-11 yet?

Re:Digg Wins

Digg is great for learning of new technologies. Slashdot is great because of its discussions. They are complementary websites. I don't see digg replacing slashdot anytime soon, yet, I am glad that digg is there to fill my need of having "as many tech news as possible" available.

Trying to have good discussions in Digg is futile because of its moderation system. And whenever discussion worthy news are available they are quickly buried by ten articles of what someone somewhere might have said about the color of the new Nintendo console.

Cheers,
Adolfo

Re:That's pretty slow

At 100Km/hour, a truck would require 1.6 hours * 60^2 seconds/hour = 5,760 seconds to travel 160 kilometers. At 60 DVDs/second, the truck would have to be carrying 5760*60 = 345,600 DVDs to have equivalent bandwidth. A typical DVD in a case is 14cm wide, 19cm tall, and 1.5cm thick, for a total volume of 399 cm^3 (lets round to 400cm^3). Therefore, the truck would have to have a cargo volume of 400cm^3 * 345,600 = 138,240,000cm^3, or 138.24m^3.

Now, typical intermodal containers (as used on big rig trucks) are 8.5' by 8.5' by 40', or 2890ft^3. Converted to metric, this is about 82m^3, which is less than the 138.24m^3 required.

In other words, no, a truck full of DVDs is NOT faster than this connection!*

*unless you put the DVDs on spindles instead of in cases.

Re:That's pretty slow

The 11,520,000 ms ping times also might interfere with some applications.

Re:That's pretty slow

Where's the moderation for "+5 for extreme nerdiness, minus several million because you really should get out more"? /says someone who's closing in on 4000 posts to slashdot...

Re:who cares?

This technology enables many more people to be simultaneously stuck at 4Mbps down/256Kbps up.

Re:who cares?

Heh. My ISP recently offered that an extra A$10/month would see my speed upgraded from 8Mbps/128Kbps (Though I tended to see about 10Mbps) to 17Mbps/256Kbps - I now regularly download 2 megabytes a second. Wee.

Re:in libraries of congress please

Research scientists are always ahead of the "real world". This has always been, always will be. You can view their work as creating ideas, innovations and technologies. Once these ideas have been published, it is the industry's work to pick them up and transform them into something commercially usable. Yes, there is a lot of research projects that can be viewed as useless, but, you should see it as a brainstorming of new technologies. Not all will end up in something revolutionary, but it may incite new ideas and/or bring new products or ways of doing things in the "real world".