GUC Joins NVIDIA NVLink Fusion Ecosystem to Drive Seamless XPU Integration
September 3, 2025 -- GUC joined the NVIDIA NVLink Fusion Ecosystem. Our purpose is to help our customers integrate their custom CPUs and GPUs into the NVIDIA MGX rack-scale infrastructure. We support porting customer needed IP to TSMC process nodes and to integrate and validate the IPs in customers' ASICs.
Two use cases:
- Integration of NVLink chiplet to custom XPU packages and corresponding UCIe and protocol IPs into main XPU die, to enable custom XPUs with NVLink interfaces
- Integration of NVLink-C2C into custom CPU to enable communication with NVIDIA GPUs
We also collaborate with a key Fusion Ecosystem partner Synopsys on integration of their IPs into customers' XPUs. NVIDIA built the most energy and latency efficient scale-up infrastructure, validated and in volume production. Our target is to enable seamless integration of our customers' XPUs with NVLink Fusion.
GUC provides full-spectrum ASIC design and manufacturing services, integrating advanced process technologies, packaging solutions, and silicon-proven IP to help customers accelerate product innovation and time-to-market.
🔍 Explore GUC's comprehensive capabilities — from turnkey ASIC solutions and high-performance IP to proven success across real-world applications.
👉 Solutions | 👉 IP Portfolio | 👉 Success Story
📩 Interested in collaboration or technical discussions? We're here to support your next innovation.
Related Chiplet
- DPIQ Tx PICs
- IMDD Tx PICs
- Near-Packaged Optics (NPO) Chiplet Solution
- High Performance Droplet
- Interconnect Chiplet
Related News
- Alchip Among the First with NVIDIA NVLink Fusion Design Infrastructure
- French AI chiplet boost for open source Nvidia accelerator
- Leaders in Semiconductors, Packaging, IP Suppliers, Foundries, and Cloud Service Providers Join Forces to Standardize Chiplet Ecosystem
- GUC Taped Out UCIe 32G IP using TSMC's 3nm and CoWoS Technology
Latest News
- Avicena Launches the World’s First microLED Optical Interconnect Evaluation Kit for AI Infrastructure Innovators
- Lightmatter Achieves Record 1.6 Tbps Per Fiber to Accelerate AI Optical Interconnect
- Arm Positions Neoverse for AI and Telco Networks at MWC
- NVIDIA Compute Architecture Paves the Way for Scale-Up Optical Interconnects; CPO Penetration in AI Data Centers Expected to Rise Steadily
- CEA-Leti and NcodiN Partner to Industrialize 300 mm Silicon Photonics for Bandwidth-Hungry AI Interconnects