JNCASR terahertz silicon chip reaches 100 Gbit/s in 6G research

By Daniel Okafor

Researchers at the Jawaharlal Nehru Centre for Advanced Scientific Research (JNCASR) developed a silicon chip on 4 November 2026 that transmitted data at 100 Gbit/s, equivalent to 12.5 gigabytes per second. The chip lets terahertz waves pass through engineered barriers with minimal signal loss. It targets compact circuits for future 6G networks, and Advanced Materials published the research.

What did JNCASR demonstrate?

The design addresses a central terahertz device problem: high signal loss. Terahertz frequencies span 0.1–10 THz and offer wide bandwidth, but signal loss has complicated equipment development. JNCASR's reported 100 Gbit/s result combines laboratory throughput with low-loss passage through engineered barriers, rather than reporting a data rate alone.

JNCASR plans to tune the barriers electrically, thermally or mechanically. Those three controls would let devices redirect signals between paths on demand. The development has no announced production date, package size, power consumption figure or operator trial, so it remains a research result rather than a deployable network component.

Why does terahertz remain difficult to deploy?

Terahertz radiation occupies the transition region between microwave and far-infrared radiation. Atmospheric gases strongly absorb these waves, attenuating most of their energy within a few metres. That physical limit constrains long-distance terrestrial radio links, and the research reports no field distance or coverage figure for the 100 Gbit/s chip.

The 6G frequency bands remain undecided. Researchers have identified 100 GHz–3 THz as promising because it contains wide swathes of unused spectrum, but that range remains a candidate rather than a settled 6G allocation. Spectrum policy therefore sits ahead of any operator rollout based on these chips.

What did related university chips achieve?

A 2024 University of Adelaide team led by Withawat Withayachumnankul developed an ultra-wideband integrated terahertz polarisation (de)multiplexer on a substrateless silicon base. The team included Weijie Gao, now at Osaka University. Testing covered the 220–330 GHz J-band, where the device carried multiple data streams in one frequency band and effectively doubled capacity.

The Adelaide team anticipated commercial prototypes and products within a decade. That is a research-team forecast, not evidence of a signed vendor deal or operator deployment. Among the three efforts, it provides the only explicit commercialization timeline, and even that timeline reaches products rather than a named network launch.

In December 2023, Nanyang Technological University researchers led by Ranjan Singh developed two chips to address signal scattering and interference. One transmitted at up to 160 Gbit/s, and light could switch it. The other isolated two channels, supporting real-time high-definition video at 1.5 Gbit/s on one channel and data transmission at 40 Gbit/s on the other.

The 160 Gbit/s figure exceeds JNCASR's 100 Gbit/s laboratory result, but the chips perform different functions and share no stated test conditions. Operators should not treat either rate as a settled 6G air-interface target or commercial network throughput. The measurements show component potential, not deployed capacity.

What changes for telecom operators?

For equipment and chip developers, the JNCASR barrier offers three tuning routes for future compact terahertz circuits. For operators, atmospheric absorption and unsettled 6G spectrum policy remain more immediate constraints than the laboratory data rate. No base-station product, operator contract, subscriber figure, deployment schedule or vendor performance claim appears in the published results.

The next milestones are electrically, thermally or mechanically tunable versions of the JNCASR barrier and the Adelaide team's hoped-for prototypes within a decade. Neither milestone supplies a 6G launch schedule: spectrum policy must settle the bands, and researchers must still show how laboratory throughput performs over practical terrestrial radio paths.