Alcatel-Lucent sends data 1600 miles at 16.4Tbps

Sure, the researchers at Alcatel-Lucent have already sent data 50 miles at 25.6Tbps, but it looks like now they're going for distance rather than speed: the company announced yesterday that it's now pumped bits over a 1584-mile long link at 16.4Tbps. Sure, that's slightly slower than the record, but being able to firehose bits at distances like that is even more impressive, if you ask us (you didn't). The core tech is essentially the same as used in the earlier speed record: bundling several 100Gbps optical signals at different wavelengths into one multiplexed transmission, shooting it down fiber, and splitting it up at the end. This latest test used 164 different channels and updated transmitters and multiplexers to hit the record -- which is fine and all, but guys, if you're not using that old school 25.6Tbps gear anymore we know a few people who are interested.

[Via Slashdot]

[eggothewaffle @ Feb 29th 2008 3:40PM]

Does anyone do proofreading anymore? :P

"This latest test used 164 different channels and updated transmitters and multiplexers to hit the record -- which is fine and all, but guys, if you're NOT using that old school 25.6Tbps gear anymore we know a few people who are interested."

Or am I going crazy?

[The Lord Dragon @ Feb 29th 2008 3:44PM]

Nope. No one proofreads anymore.

But you're still going crazy.

[pundit @ Feb 29th 2008 3:44PM]

Yeah. That made me pause for a second too. It's pretty sweet that they've got this (ridiculously expensive) tech in development though.

[Raheem @ Feb 29th 2008 4:17PM]

Wait... what?

[justincase @ Feb 29th 2008 4:20PM]

I asked..

[Shades @ Feb 29th 2008 4:23PM]

Where's the proofreading problem? They used fewer channels on the second experiment, but all of the math works.

[pundit @ Feb 29th 2008 5:01PM]

Before Engadget quietly edited the post, they were missing the "not" in the sentence "if you're NOT using that old school ..." they quietly fixed the error though.

[a ham sandwich @ Feb 29th 2008 5:01PM]

since that picture is so small, at first i thought it was one of those old school room-sized computers. lol.


http://en.wikipedia.org/wiki/Eniac

[ShadowMaker @ Feb 29th 2008 5:04PM]

How much porn would that be a second?

[Nathan @ Feb 29th 2008 5:07PM]

16.4 Terra-boners / second

[jon @ Feb 29th 2008 5:05PM]

actually it makes sense that the throughput on the longer distance would be less because of signal segregation.
"The researchers used 164 wavelength-division multiplexed channels modulated at 100 Gbps in the effort."
so that would be one optical strand, how many can they fit into a bundle like the undersea cables?

[Clark H. @ Feb 29th 2008 7:15PM]

Shit, it takes me about five days to download a DVD!!
Just think about it: I could download a whole Blockbuster in a few minutes with cables like that!!

O_O

[SimonRichards @ Feb 29th 2008 9:36PM]

5.45 Seconds according to my calculations, which are probably wrong anyway. :D

[RandomCake @ Mar 1st 2008 7:29AM]

Unless I'm missing something, your calculations are wrong :S The transfer rate would give you over 2TB (2,000GB) of data a second (obviously that's going to be shared between a few users :P) but still, you'd get your movie in under a second, even if it's a direct bit for bit copy of a Blu-ray Disk ;)

[ev01 @ Mar 1st 2008 8:53AM]

It would be even sweeter if we had computer disks and a bus that could write at 2TB per second. Your hardware now becomes the bottleneck.

[Tony Rayo @ Mar 1st 2008 5:32PM]

Yeah I'm wondering about that too. Did they use SSD's for this transfer? Or directly into RAMBUS? Can either of those technologies even operate at that speed? I'm more curious in how they stored the data than how they transfered it. If data was just shot out and read with an optical version of a speedometer, then that's pointless IMO.