from the kilo-mega-giga-tera-peta...wow-that's-*FAST* dept.
Record 1.84 Petabit/s Data Transfer Achieved With Photonic Chip, Fiber Optic Cable
Scientists from the Technical University of Denmark in Copenhagen have achieved 1.84 petabits per second data transfers using a single photonic chip connected via a single optical fiber cable. The feat was accomplished over a distance of 7.9 km (4.9 miles). For some perspective regarding this achievement, at any time of day, the average internet bandwidth being used by the whole world's population is estimated to be about 1 petabit/s.
[...] Firstly, the data stream used in the trial was split into 37 lines, with each one sent down a different optical thread in the cable. Each of the 37 data lines were split into 223 data chunks corresponding to zones of the optical spectrum. What this allowed is for creating a "frequency comb" where data was transmitted in different colors at the same time, without interfering with other streams. In other words a "massively parallel space-and-wavelength multiplexed data transmission" system was created. Of course, this splitting, and re-splitting massively increased the potential data throughput supported by a fiber optic cable.
[...] In action, the photonic chip splits a single laser into many frequencies and some processing is required to encode light data for each of the 37 data optical fiber streams. A refined fully capable optical processing device should be possible to build at approximately the size of a match box, according to Jørgensen. This is a similar size to current single color laser transmission devices used by the telecoms industry.
Also at Notebookcheck.
Journal Reference:
A. A. Jørgensen, D. Kong, M. R. Henriksen, et al.
Petabit-per-second data transmission using a chip-scale microcomb ring resonator source (DOI: 10.1038/s41566-022-01082-z) (DX)
(Score: 3, Insightful) by legont on Sunday October 23 2022, @12:49AM (5 children)
I think generating that much data - twice the whole internet - is a bigger achievement than the cable.
"Wealth is the relentless enemy of understanding" - John Kenneth Galbraith.
(Score: 2) by takyon on Sunday October 23 2022, @12:52AM (4 children)
[SIG] 10/28/2017: Soylent Upgrade v14 [soylentnews.org]
(Score: 2) by legont on Sunday October 23 2022, @01:36AM (3 children)
I done those kind of tests in the office - not that scale obviously. It is hard to generate a lot of data on the fly. If data was pre generated, it is even harder to read it fast enough. I don't know much about fiber so I guess I accept it, but I'd really wanted to see how they got the data. Perhaps some analog device? Even some quantum thing? Look, it's twice the size of the whole traffic.
I hope they did not load one channel and said that many of them would handle that many multiplied. That'd be fishy.
"Wealth is the relentless enemy of understanding" - John Kenneth Galbraith.
(Score: 3, Informative) by BigJ on Sunday October 23 2022, @02:26AM (2 children)
PRBS algorithm https://en.m.wikipedia.org/wiki/Pseudorandom_binary_sequence [wikipedia.org] generates data on chip at line speed. Since it's deterministic, you can check for transmission errors.
(Score: 2) by legont on Sunday October 23 2022, @02:57AM (1 child)
That's a variation of the original pseudorandom multiplication algorithm which in FORTRAN takes just one cycle. Way not fast enough though.
"Wealth is the relentless enemy of understanding" - John Kenneth Galbraith.
(Score: 4, Interesting) by BigJ on Sunday October 23 2022, @04:53AM
Plenty fast when it's performed in an ASIC. It's just a linear feedback shift register. Each bit clock shifts out a new value. No need for a general purpose cpu to generate data. Each WDM channel would have its own PRBS logic. 1.84pb/s / 37 / 223 = 223gb/s per WDM channel. That's Ethernet speeds. Current silicon logic would have no problems generating test patterns at that line rate.
(Score: 2) by inertnet on Sunday October 23 2022, @11:12AM (1 child)
That's 121 gigabit per data stream. I wonder how much heat that chip generates.
The journal says they think a total of 100 petabit per second for a single chip is possible:
(Score: 2) by takyon on Sunday October 23 2022, @12:47PM
The paper isn't available in the usual places right now. I'll check back later.
If those speeds aren't making it out of the lab anytime soon, maybe a more basic/modest version will.
[SIG] 10/28/2017: Soylent Upgrade v14 [soylentnews.org]
(Score: 2, Funny) by Anonymous Coward on Sunday October 23 2022, @01:33PM
Just what I need to hook up to my beowolf cluster!
(Score: 2) by krishnoid on Sunday October 23 2022, @05:57PM
Transmission protocols have been moving away from parallel, and towards serial for a couple decades, and this is a step backwards. They should just get faster processors and with a few tweaks [youtu.be] they could clock the photons out faster without those synchronization issues on the receiving end.