Chinese researchers say they achieved 3.2 petabits per second (Pb/s) of real-time, bidirectional optical transmission through a single 24-core fiber. The collaboration involved China Information and Communication Technologies Group’s optical-communications laboratory, Pengcheng Laboratory and Fiberhome Fujikura. A October 26 Science and Technology Daily report describes a 10.3-kilometer experiment, while Wuhan’s government website reported the result on October 23.
A 28% Increase From the Earlier Experiment
The same research team documented 2.5 Pb/s in a March 2026 Optical Fiber Communication Conference paper. Increasing the reported aggregate rate to 3.2 Pb/s adds 0.7 Pb/s, a 28% capacity gain. That is not a 28% increase in household internet speed. The new figure is a research-team record claim rather than a separately certified commercial-network result.
How 24 Cores Carry More Information
Instead of relying on one optical path, multicore fiber contains multiple light-carrying cores inside one strand. The researchers combined space-division multiplexing across 24 cores with wavelength channels spanning the S, C and L bands. They reported using commercially compatible coherent optics while managing interference between cores, known as crosstalk.
That strategy addresses a practical density problem. Our earlier coverage of Corning’s four-core fiber shows why increasing capacity inside existing cable pathways matters as AI facilities add optical connections. Corning’s video above explains the underlying technology; it does not independently verify China’s latest claim.
Why Data Centers Are Watching
Petabit-scale aggregate fiber capacity could eventually help link large computing campuses, carriers and cloud regions. BitcoinVersus.Tech also covered Meta’s planned 1 Pb/s Petal subsea project, which uses a different multicore architecture over a much longer proposed route. These figures describe separate systems and should not be treated as equivalent field results.
The Meta announcement above is an authentic, separate industry example. The Chinese experiment establishes neither an immediate consumer broadband upgrade nor a deployment timetable. Engineers still need published performance data, including transmission errors, operating stability and long-term deployment costs. Our optical link-budget lesson explains why reach and signal margins matter alongside headline bandwidth.
What Comes Next
Science and Technology Daily reported that the 3.2 Pb/s work was submitted by invitation to the Journal of Lightwave Technology. Publication of detailed measurements and independent assessment will help establish how reproducible and deployable this claimed record is.
Editor’s Note: We volunteer daily to ensure the credibility of the information on this platform is Verifiably True. Please support independent reporting and research with a 5,000-sat donation by scanning the QR code below.
Scan the donation QR code from another device or a larger screen if your phone cannot scan its own display. This report is informational and does not provide financial advice.
Leave a Reply