en.Wedoany.com Reported - Recently, Keysight Technologies launched an end-to-end electrical-optical-electrical (EOE) simulation solution for data center and Ethernet designs. Integrated into ADS 2026, this solution enables engineers to simulate the complete link from electrical signals to optical signals and back to electrical signals within the same design environment, serving high-speed optical interconnects, AI data centers, and high-performance computing network designs.
The core of this solution is integrating design workflows previously scattered across circuit simulation and photonic simulation tools into a unified workflow. AI infrastructure is driving 800Gbps, 1.6Tbps, and even higher-rate optical links into data center interconnect scenarios. The interactions between CPUs, GPUs, high-speed SerDes interfaces, optical modules, photonic integrated circuits, and electrical receivers are becoming increasingly complex. Traditional workflows typically model the electrical and optical domains separately, with engineering teams manually stitching results together, which can easily miss cross-domain effects, such as signal integrity issues arising from the combined behavior of high-speed SerDes digital channels and photonic chips. By introducing EOE simulation capabilities in ADS 2026, Keysight allows engineers to simultaneously evaluate the system-level performance of electrical channels, optical devices, modulators, receivers, and multi-wavelength links before hardware implementation, enabling earlier identification of performance bottlenecks and architectural trade-offs.
This solution combines Keysight's high-speed digital workflow with Keysight Photonic Designer, supporting complete signal path simulation, bidirectional optical link simulation, multi-wavelength nonlinear effect assessment, and cross-electrical-optical domain noise modeling.
For data center and Ethernet design teams, end-to-end EOE simulation means development risks can be shifted forward for control. High-speed optical interconnects are moving from traditional module-level verification to a stage that emphasizes collaborative design of chips, packages, photonic devices, and systems. As optical link rates increase, multi-channel wavelength division multiplexing, modulator bias, large-signal nonlinearities, return reflections, noise accumulation, and system-level signal quality all impact final performance. If these issues are only discovered during the sample or prototype stage, design modification costs increase significantly. The unified simulation environment provided by ADS 2026 helps engineering teams complete more verification in advance during architecture definition, link budget analysis, optical-electrical interface design, and high-speed standard evaluation, while reducing errors caused by data conversion between different tools.
This release also reflects the deep extension of test and measurement and EDA software into the AI infrastructure design chain. The high-speed interconnect capability of data center networks has become a key constraint for the expansion of AI training clusters and inference systems. Optical interconnect design is no longer just an internal issue for optical module vendors but is closely related to GPU clusters, switching chips, packaging, rack networking, and overall data center architecture. By incorporating EOE simulation into ADS 2026, Keysight expands its software capabilities in photonic integration, high-speed digital, semiconductor, and data center interconnect design, while also providing engineers with a continuous design path from system-level to component-level optimization.
As AI computing infrastructure evolves towards higher rates, higher density, and lower power consumption, the importance of collaborative electrical and optical domain design will continue to rise. The subsequent effectiveness of this simulation solution launched by Keysight will depend on model accuracy, PDK support scope, the completeness of the photonic device library, and verification feedback from engineering teams in real projects. For the high-speed Ethernet and AI data center supply chain, identifying cross-domain issues early in the simulation phase will become a key step in shortening development cycles and improving system reliability.
This article is compiled by Wedoany. All AI citations must indicate the source as "Wedoany". If there is any infringement or other issues, please notify us promptly, and we will modify or delete it accordingly. Email: news@wedoany.com
