Keysight Technologies has released SOS Enterprise, an upgraded version of its engineering data management platform aimed at supporting enterprise-scale AI adoption within semiconductor design.

This new platform builds on the existing SOS Core software, providing organisations with automated engineering data governance and traceability capabilities required for deploying AI at scale.

In the current landscape, semiconductor and electronics engineers encounter rising design complexity and rapidly expanding datasets, issues that intensify across distributed, multi-site environments.

According to Keysight Technologies, rapid advancements in semiconductor and electronics development are driving increased design complexity and larger volumes of engineering data, especially for organisations managing operations across several locations.

Data silos across fragmented tools and file systems often result in significant productivity loss due to manual search processes, inconsistent data versions, and additional overhead resolving integrity and correlation conflicts.

Keysight Technologies points out that insufficient traceability not only complicates compliance but also impedes the reliable operation of machine learning solutions, which rely on consistent and well-organised data.

SOS Enterprise addresses these concerns by establishing a unified system of record across geographically dispersed teams, said the company. The platform enables clear data lineage, automated compliance tracking, and standardised management of engineering information across diverse regulatory contexts.

By storing design files and verification data as versioned and traceable assets, SOS Enterprise aims to integrate seamlessly with AI-based design and verification pipelines.

The company stated that the platform delivers audit trails, role-based access control, and software bill of materials tracking while maintaining current engineering workloads. This supports compliance requirements in industries like aerospace, defence, and automotive.

The software enforces granular access controls using fine-tuned permissions, geofencing measures, and compliance-driven restrictions.

Traceability is maintained throughout the design lifecycle by automating workflows and tracking the origin and relationships of engineering data. This approach enables engineering teams to prepare their datasets for scalable AI integration while allowing for validated IP reuse via a centralised record.

Keysight Technologies data and IP management general manager Simon Rance said: “Engineering organisations tell us they’re ready to adopt AI, but their data isn’t. When design files, verification results, and IP libraries are scattered across disconnected systems, AI tools can’t deliver trustworthy outcomes.

“We built SOS Enterprise to solve that foundational problem by automating the governance and traceability work that’s currently done manually, making their engineering data AI-ready so teams can focus on innovation rather than data management.

“Early customer deployments have demonstrated significant improvements in IP reuse, faster project orchestration, and reduced operational overhead.”

Last month, Keysight Technologies entered a collaboration with Point2 Technology to validate next-generation multi-terabit interconnects for AI and machine learning data centres.

Utilising test solutions such as the M8050A High-Performance Bit Error Ratio Tester (BERT) and DCA-X Sampling Oscilloscopes, Keysight Technologies is working with Point2 Technology to assess its e-Tube technology’s suitability for hyperscale AI infrastructure.

The e-Tube solution employs radio frequency (RF) transmission over plastic waveguide to target bandwidth bottlenecks typical in accelerator processing unit (xPU) cluster scaling while aiming for competitive cost structures.

Keysight Technologies is providing a validation framework to meet hyperscaler reliability and performance standards. The collaboration also gives the industry early access to research 3.2T interfaces by leveraging Keysight Technologies’ 448Gbps electrical signal generation technologies paired with advanced real-time analysis instrumentation.