MaxLinear Unveils Rushmore™: Low-Power 1.6T PAM4 DSP for AI/ML and Data Center Networks

In This Article:

MaxLinear Rushmore Low Power 1.6T PAM4 DSP
MaxLinear Rushmore Low Power 1.6T PAM4 DSP
  • Rushmore integrates our latest advancements in DSP technology to address power demands and drive performance in the rapidly expanding AI/ML space

CARLSBAD, Calif., March 31, 2025--(BUSINESS WIRE)--MaxLinear, Inc. (Nasdaq: MXL), a leading provider of high-speed interconnect ICs enabling data center, metro, and wireless transport networks, is excited to announce the sampling and commercial availability of its 1.6T (200G/lane) PAM4 SERDES (Serializer/Deserializer) and DSP, Rushmore. Built on Samsung’s leading-edge CMOS technology, this DSP delivers a substantial boost in performance, power efficiency and density, enabling 1.6T optics and active copper cables, and paves the way for the next generation of high-speed, efficient networking.

Rushmore leverages three generations of MaxLinear's advanced DSP technology, encompassing 100G to 800G devices. This heritage of innovation ensures exceptional performance and reliability. Coupled with robust firmware, Rushmore delivers superior power-efficiency and functionality, while meeting the stringent performance requirements of 1.6TB highspeed optical connectivity solutions.

At OFC 2025 in San Francisco, MaxLinear will present live demonstrations of the new 1.6T Rushmore DSP, highlighting its performance and capabilities in enabling < 25W optical modules. "Rushmore was specifically optimized to address the next generation of 1.6T interconnects, featuring low power consumption, excellent link performance, and enhanced features to accelerate deployments and monitor link health," said Kishore Seendripu, Chief Executive Officer at MaxLinear.

Key Features and Benefits of Rushmore:

  • High-Speed: Capable of up to 200G per lane, compatible with Ethernet and InfiniBand® protocols.

  • Power Efficiency: Designed to support the next generation of <25W DR/FR optical modules, adhering to OIF, IEEE, and IBTA standards.

  • Low Latency Architecture: Optimized for the low latency demands of AI/ML clusters.

  • Low Power Modes: Includes special features and modes to optimize link power budgets.

  • Advanced Diagnostics: Equipped with MPI detection and localization, full 8x8 crossbar, and pre-FEC BER monitor for comprehensive link health monitoring.

  • Automated DPD (Digital Pre-Distortion): Optimizes transmitter performance across all laser technologies, including EMLs and silicon photonics.

  • Integrated Driver Option: Offers an integrated driver optimized for high-output swing and reduced system power consumption.

  • Best-in-Class DSP + TIA: Co-designed to partner with MaxLinear’s Washington TIA to provide the best performance on the market.