The increasing complexity of advanced semiconductor designs is forcing a shift toward continuous, real-time silicon visibility to manage reliability and power. proteanTecs and Global Unichip Corporation (GUC) have expanded their strategic collaboration to integrate proteanTecs’ health and performance monitoring solutions directly into GUC’s custom ASIC designs. This move aims to provide GUC customers with in-mission power reduction, performance optimization, and enhanced reliability. By embedding these monitoring systems into the compute fabric, the companies intend to address critical demands for resilience in data centers, automotive, and AI infrastructure.
Expanding Monitoring into the Compute Fabric
Building on previous work involving high bandwidth die-to-die interconnects, HBM subsystems, and advanced packaging, the two companies are now broadening the deployment of proteanTecs solutions into the core compute fabric itself. This expansion allows GUC to offer proteanTecs as a pre-validated solution for its ASIC customers, which the companies claim will reduce time-to-market through seamless integration. By embedding the proteanTecs hardware monitoring system into GUC’s custom designs, customers gain access to real-time telemetry. This data is intended to support advanced diagnostics, predictive maintenance, and fleet management once the silicon is deployed in the field. The collaboration targets high-growth sectors including networking, AI infrastructure, and physical AI, where tighter guard-bands and lower power consumption are increasingly vital for maintaining system stability under heavy, real-world workloads.
Addressing Reliability in Advanced Compute Designs
As process technology and system integration push closer to physical limits, GUC CTO Igor Elkanovich notes that customers require better predictability regarding how silicon behaves in real-world systems. The expanded partnership seeks to provide this predictability by allowing users to control and optimize long-term performance via runtime monitoring. The integrated solution is designed to support the entire lifecycle of advanced compute designs, spanning from initial silicon bring-up through production and into active field operation. This lifecycle approach is positioned to help enterprises manage the complexities of heterogeneous integration and high-performance compute IP. By making deep, in-silicon visibility an integral part of the design phase rather than an afterthought, the collaboration aims to provide a comprehensive monitoring foundation for the next generation of complex, large-scale semiconductor deployments.
Key Takeaways
- GUC will provide proteanTecs’ health and performance monitoring solutions to its ASIC customers to enable in-mission power and performance optimization.
- The collaboration expands beyond HBM subsystems and interconnects to include the integration of monitoring systems into the compute fabric itself.
- The solution targets critical infrastructure sectors, specifically data centers, networking, automotive, AI infrastructure, and physical AI.
TechInsyte's Take
In our view, this expansion signals a fundamental shift in how enterprise-grade silicon must be managed. As AI and data center workloads push hardware to its thermal and electrical limits, the traditional "set and forget" model of semiconductor deployment is becoming a liability. By embedding proteanTecs' monitoring directly into GUC's ASIC fabric, the companies are moving toward a model of "active silicon," where telemetry drives real-time adjustments. This is not just about detecting failure; it is about managing the degradation and efficiency of high-value assets in real-time. For CIOs, this suggests that future hardware procurement will increasingly depend on the depth of integrated telemetry available.
Questions & Answers
How does this collaboration impact the time-to-market for GUC customers?
The collaboration provides a tested and pre-validated solution, which GUC states will enable reduced time-to-market through more seamless integration of monitoring capabilities into custom ASIC designs.
Which specific industry applications are targeted by this expanded monitoring solution?
The companies are targeting applications that require high resilience and lower power, specifically data centers, networking, automotive, AI infrastructure, and physical AI.
What technical capabilities does the embedded monitoring system provide?
The system provides real-time power, performance, and reliability optimization, alongside rich telemetry intended for advanced diagnostics, predictive maintenance, and fleet management in the field.
Beyond the compute fabric, what other technologies were previously covered in this partnership?
Prior to this expansion, the collaboration focused on monitoring solutions for high bandwidth die-to-die interconnects, HBM subsystems, and advanced packaging.
Source: proteanTecs
