10.2 Terabit/s over 29 km on optical fibre

[tezunegari]

Source

(openPR) - In a new world premier, members of Berlin’s Fraunhofer Heinrich Hertz Institute in collaboration with the Technical University of Denmark in Copenhagen have succeeded in generating a serial data rate of 10.2 Terabit per second and transmitting it over a 29 km long fiber optic link. This breakthrough was achieved by an even more rapid and comprehensive modulation of the light signal transported in the fiber optic cable on a single optical carrier or wavelength. This new record puts the previous world record held by the Heinrich Hertz Institute of 2.65 Terabit per second clearly in the shade. On March 10th , 2011, the Fraunhofer HHI scientists will be presenting their new record for the first time to audiences at the Optical Fiber Communications Conference (OFC 2011) in Los Angeles.

Fraunhofer HHI’s new record is based on two innovations. On the one hand, the data transmission pulse repetition rate is increased by switching on and off the flashes transmitting the data in the fiber optic cable much more rapidly and with shorter breaks than ever before. In other words, every 800 femoseconds researchers sent an optical pulse of just 300 femoseconds pulse duration down the optical cable, giving a pulse repetition rate of 1.28 THz – with 32 times more repetition than in commercial 40 Gbit/s transmission systems. This means that data can be time-interleaved with very high density. On the other hand, researchers increased the amount of information transmitted in each flash by modulating not just the amplitude but the optical phase of the electromagnetic light wave as well. 16-QAM modulation, used for the first time on such an ultra-rapid sequence of optical pulses, enabled coding of 4 bits per pulse which means that an extra four times as much information could be transmitted than with commercial systems.

In today’s information and communication age, fiber optical networks are the basis for modern high capacity data and communication transmission systems. Weathering all economic recessions, traffic in telecommunications networks increases steadily by over 50% each year. To meet such relentless demand, it is vital to increase the capacity of fiber optical networks and to make better use of fiber optical links if we are going to enable future transmission of modern services and high resolution video. The technology used in the world record experiment was developed by Fraunhofer HHI in the “Optical Technologies” program (13N9356) funded by the Federal Ministry of Education and Research (BMBF).

Scientists at Fraunhofer HHI will be making the first public presentation of their work on 10 March 2011 at the Optical Fiber Communications Conference (OFC 2011) in Los Angeles. The OFC is the world’s largest global conference on optical communications technology and attracts over 9,000 attendees each year.

The Fraunhofer Heinrich Hertz Institute is a global leader in the development of mobile and fixed broadband communication networks and multimedia systems. From photonic components and systems and fiber optic sensor systems through to high-speed hardware architectures, the Heinrich Hertz Institute works together with international partners from research and enterprise and for global markets on developing the infrastructures for the future Gigabit Society. At the same time it also develops future applications for broadband networks. Key focal areas of research are 3D TV, 3D displays, HDTV, gesture-controlled human-machine interaction, image signal processing and transmission, and interactive use of media.

I wonder how long it will take until this sees real world implementation...

[Eternal_Freedom]

It will come into common use when we have a realistic need to send 10+ TB amounts of data in one go.

[Mr Bean]

It will come into common use when we have a realistic need to send 10+ TB amounts of data in one go.

You mean now?
Content expands to fill the void not the other way around. If we had 10 Terabyte connections we would quickly come up with programs and devices that would benefit from that much raw throughput. We might see a partial return to dump terminals if your Notebook can download the contents of your entire home computer on the fly. It also makes distribution of movies much eaiser. No need to send physical prints at all or burn blue-ray's when the entire movie can be transmitted in under a second.

[cosmicalstorm]

Does work like this have any bearing on how much information can be moved around inside one computer?

[General Zod]

Does work like this have any bearing on how much information can be moved around inside one computer?

What do you mean 'moved around inside'? You're still going to have limits depending on hard disk size and type.

[Starglider]

Does work like this have any bearing on how much information can be moved around inside one computer?

Eventually, yes. The problem is that carrier-grade optical WAN transceivers cost tens to hundreds of thousands of dollars. The acceptable cost of transceivers that can go on a motherboard is a few cents each. Intel just removed the optical component of their next-gen interface (Lightpeak, making all those Apple ads that go on about it even more vapid than usual) because the costs were still too high; and that's in an application that could probably tolerate 50 cent transceivers. Thus so far lots of talk and R&D about optical on-pcb and on-chip, but nothing rolled out.

[Uraniun235]

It will come into common use when we have a realistic need to send 10+ TB amounts of data in one go.

This will probably be most useful in transoceanic communications cables, whose capacities are in some cases already measured in terabits per second.

[cosmicalstorm]

Does work like this have any bearing on how much information can be moved around inside one computer?

Eventually, yes. The problem is that carrier-grade optical WAN transceivers cost tens to hundreds of thousands of dollars. The acceptable cost of transceivers that can go on a motherboard is a few cents each. Intel just removed the optical component of their next-gen interface (Lightpeak, making all those Apple ads that go on about it even more vapid than usual) because the costs were still too high; and that's in an application that could probably tolerate 50 cent transceivers. Thus so far lots of talk and R&D about optical on-pcb and on-chip, but nothing rolled out.

Thanks, that was enlightening.