Synopsys, MediaTek, Marvell, and Others Embrace NVIDIA's NVLink Fusion for AI Workloads
May 19, 2025
News
COMPUTEX—NVIDIA's NVIDIA NVLink Fusion, new silicon designed to enable the build of semi-custom AI infrastructure, has been adopted by MediaTek, Marvell, Alchip Technologies, Astera Labs, Synopsys, and Cadence, enabling custom silicon scale-up to support workloads for model training and agentic AI inference.
Using NVLink Fusion, Fujitsu and Qualcomm Technologies CPUs can also be integrated with NVIDIA GPUs to build high-performance NVIDIA AI factories.
Per the company, NVLink Fusion equips cloud providers with a path to scale out AI factories to millions of GPUs, using any ASIC, NVIDIA’s rack-scale systems, and the NVIDIA end-to-end networking platform, which delivers up to 800Gb/s of throughput and features NVIDIA ConnectX®-8 SuperNICs, NVIDIA Spectrum-X Ethernet, and NVIDIA Quantum-X800 InfiniBand switches, with co-packaged optics available soon.
Using NVLink Fusion, hyperscalers can work with the NVIDIA partner ecosystem to integrate NVIDIA rack-scale solutions for deployment in data center infrastructure.
NVLink Fusion also enables AI innovators like Fujitsu and Qualcomm Technologies to each couple their custom CPUs with NVIDIA GPUs in a rack-scale architecture to boost AI performance.
To maximize AI factory throughput and performance in a power-efficient way, the fifth-generation NVIDIA NVLink platform includes NVIDIA GB200 NVL72 and GB300 NVL72, compute-dense racks that provide a total bandwidth of 1.8 TB/s per GPU — 14x faster than PCIe Gen5.
Leading hyperscalers are already deploying NVIDIA NVLink full-rack solutions and can speed time to availability by standardizing their heterogeneous silicon data centers on the NVIDIA rack architecture with NVLink Fusion.
AI factories connected with NVIDIA NVLink Fusion are powered by NVIDIA Mission Control, a unified operations and orchestration software platform that automates the management of AI data centers and workloads.
NVIDIA Mission Control is designed to improve AI factory operations, from configuring deployments to validating infrastructure to orchestrating mission-critical workloads, to helping enterprises get frontier models up and running faster.
NVIDIA NVLink Fusion silicon design services and solutions are available now from MediaTek, Marvell, Alchip, Astera Labs, Synopsys, and Cadence.
Tiera Oliver is the assistant managing editor at Embedded Computing Design. She is responsible for web content editing, product news, and story development. She also manages, edits, and develops content for ECD podcasts, including Embedded Insiders.
She utilizes her expertise in journalism and content management to oversee editorial content, coordinate with editors, and ensure high-quality output across web, print, and multimedia platforms. She manages diverse projects, assists in the production of digital magazines, and hosts company podcasts by conducting in-depth interviews with industry leaders to deliver engaging and insightful discussions.
Tiera attended Northern Arizona University, where she received her bachelor's in journalism and political science. She was also a news reporter for the student-led newspaper, The Lumberjack.
