Texas Instruments unveils power-monitoring, 800V protection and timing components for AI data centers
Редакция ЦОД:Главное
Texas Instruments has introduced three components targeting AI data-center power telemetry, 800V power distribution and high-speed networking at the OCP Global Summit in San Jose, California. The ADC34DSO75, LMI800PS and LMK6B address power-fault detection, protection and network timing at the hardware layer.
- TI introduced three components for AI data-center telemetry, power distribution and networking.
- The ADC34DSO75 samples at 3 GSPS for remote monitoring of low-voltage, high-current power rails.
- The LMI800PS is an isolated hot-swap controller for 800V power distribution.
- TI says the LMI800PS replaces a five-IC discrete design with one high-density package.
- LMK6B is a BAW oscillator for high-speed networking equipment.
Texas Instruments introduced the ADC34DSO75, LMI800PS and LMK6B at the OCP Global Summit in San Jose, California, positioning the devices for AI data-center power, telemetry and networking infrastructure.
The ADC34DSO75 is designed for continuous remote monitoring of low-voltage, high-current power rails. It samples at 3 GSPS and is intended to identify short-duration power anomalies at nanosecond scale. TI says the device provides fully remote, event-based telemetry, potentially improving visibility into transient rail disturbances that can be difficult to diagnose through local inspection.
For power distribution, the LMI800PS is an isolated hot-swap controller for 800V systems. TI says it consolidates a discrete design built from five integrated circuits into a single high-density package. The available information does not provide further technical specifications or application conditions for the controller.
The third device, LMK6B, is a bulk acoustic wave (BAW) oscillator for high-speed networking equipment. In the context of AI data centers, the three products address related hardware requirements: earlier detection of power-rail anomalies, protection within the power path, and precise clocking for network equipment.
The announcement reflects a potential shift in AI infrastructure competition toward components that improve observability and resilience directly within power and networking paths. The operational impact, commercial availability, pricing and deployment volumes were not disclosed.
As AI workloads raise requirements for power resilience, diagnostic speed and timing quality, component-level monitoring and protection are becoming more important in data-center electrical and network paths. If the stated performance is validated in operation, these products could reduce reliance on local manual diagnostics and support higher-voltage rack-power architectures.
- Data Centre Magazine / 13 октября 2026 г. · АнглийскийПервоисточник →
Материал подготовлен редакцией ЦОД: Главное на основе указанных первоисточников.