- Nvidia Integrates Samsung Foundry to Extend NVLink Fusion to Create Custom AI Chips
- NVLink Fusion allows CPUs, GPUs and accelerators to communicate seamlessly with each other.
- Intel and Fujitsu can now create processors that connect directly to Nvidia GPUs.
Nvidia is deepening its efforts to become indispensable in the artificial intelligence space by expanding its NVLink Fusion ecosystem.
After a recent collaboration with Intelwhat x86 allows processors To connect directly to Nvidia platforms, the company has engaged Samsung Foundry to help design and manufacture custom processors and XPUs.
The move, announced during the 2025 Open Compute Project (OCP) global summit in San Jose, demonstrates Nvidia's commitment to expanding its control over the entire AI computing hardware stack.
Integration of new players into NVLink Fusion
Ian Buck, Nvidia's vice president of HPC and hyperscale, explained that NVLink Fusion is an IP and chiplet solution designed to integrate processors, GPUsand accelerators are easily integrated into the MGX and OCP infrastructure.
It enables direct, high-speed communication between processors in rack-mount systems, aiming to eliminate traditional performance bottlenecks between computing components.
During the summit, Nvidia introduced several ecosystem partners, including Intel and Fujitsu, who are now capable of creating processors that interface directly with Nvidia GPUs via NVLink Fusion.
Samsung Foundry joins this list, offering a full range of services from design to custom silicon manufacturing, strengthening Nvidia's position in semiconductor manufacturing.
The collaboration between Nvidia and Samsung reflects a growing shift in the AI hardware market.
As AI workloads expand and competition intensifies, Nvidia's custom processor and XPU designs aim to ensure its technologies remain central to the next generation. data centers.
According to TechPowerUpNvidia's strategy has strict limitations.
Custom chips developed under NVLink Fusion must connect to Nvidia products, with Nvidia retaining control over communications controllers, PHY layers, and licensing of NVLink switches.
This gives Nvidia significant leverage in the ecosystem, although it also raises concerns about openness and interoperability.
This tightening integration comes due to the emergence of competitors such as OpenAI, GoogleAWS, Meta and Broadcom are developing their own chips to reduce their dependence on Nvidia hardware.
Nvidia is moving deeper into the fabric of AI infrastructure, making its technology inevitable rather than optional.
With NVLink Fusion and the addition of Samsung Foundry to its dedicated semiconductor ecosystem, the company is expanding its reach across the entire hardware stack, from silicon to data center architectures.
This reflects a broader trend among its competitors. Broadcom is diving deeper into artificial intelligence, offering dedicated accelerators for hyperscalers.
OpenAI is also reportedly developing its own chips to reduce its dependence on Nvidia GPUs.
Taken together, these developments mark a new stage of competition in artificial intelligence hardware, where control of the silicon-to-software pipeline determines who will lead the industry.
Nvidia's partnership with Samsung appears to be aimed at counteracting this by accelerating the adoption of customized solutions that can be quickly deployed at scale.
By embedding its intellectual property into broader infrastructure projects, Nvidia is positioning itself as an integral part of modern AI factories, rather than just a GPU supplier.
Despite Nvidia's contributions to the OCP open hardware initiative, its NVLink Fusion ecosystem maintains strict boundaries that favor its architecture.
While this can provide performance benefits and ecosystem consistency, it can also increase concerns about vendor lock-in.
Follow TechRadar on Google News. And add us as your preferred source to get our expert news, reviews and opinions in your feeds. Be sure to click the “Subscribe” button!
And of course you can also Follow TechRadar on TikTok for news, reviews, unboxing videos and get regular updates from us on whatsapp too much.
