Digital Event Horizon
d-Matrix, a leading AI inference chipmaker, has adopted NVIDIA's NVLink Fusion technology to accelerate its AI inference capabilities and streamline its deployment process. This strategic move marks a significant milestone in the company's efforts to provide customers with faster, lower-risk access to ultralow-latency inference solutions, and is part of NVIDIA's broader efforts to accelerate the adoption of AI and machine learning across industries.
d-Matrix partners with NVIDIA to adopt NVLink Fusion technology for its Raptor XPUs. The partnership aims to accelerate AI inference capabilities and streamline deployment. d-Matrix's goal is to provide customers with a faster, lower-risk path to deploy and scale ultralow-latency inference solutions. NVLink Fusion enables integration of custom XPUs into NVIDIA's rack-scale architecture. The technology provides high-bandwidth, low-latency connections between XPUs and CPUs. The partnership is part of NVIDIA's efforts to accelerate AI adoption across industries.
d-Matrix, a leading AI inference chipmaker, has announced its partnership with NVIDIA to adopt the company's NVLink Fusion technology for connecting its next-generation Raptor XPUs to NVIDIA's AI infrastructure platform. This strategic move marks a significant milestone in the company's efforts to accelerate its AI inference capabilities and streamline its deployment process.
According to d-Matrix's CEO, Sid Sheth, the company's primary goal is to provide customers with a faster, lower-risk path to deploy and scale ultralow-latency inference solutions. By integrating its Raptor XPUs with NVIDIA's NVLink scale-up and Spectrum-X scale-out networking, the company aims to tap into the benefits of NVIDIA's rack-scale architecture and the broader AI platform.
The adoption of NVLink Fusion is seen as a major step forward for d-Matrix, as it enables the company to integrate its custom XPUs into a broadly deployed, liquid-cooled architecture. This, in turn, will allow customers to access a faster, more reliable, and more scalable inference solution, with reduced risks and costs.
NVLink Fusion is a key component of NVIDIA's AI infrastructure platform, which spans compute, networking, storage, security, power, cooling, and software. The technology provides a high-bandwidth, low-latency connection between custom XPUs and CPUs, and can support all major CPU architectures, including Arm, x86, and RISC-V.
By adopting NVLink Fusion, d-Matrix will be able to standardize on a common rack, enabling data centers to be built once and supporting a wide range of processor types, including GPUs, CPUs, and XPUs. This will simplify the deployment process and reduce the risks associated with custom silicon development.
The company plans to use NVLink Fusion to connect its XPUs in a single high-bandwidth, low-latency scale-up domain, alongside NVIDIA GPU-based systems, such as the NVIDIA Vera Rubin NVL72. Additionally, d-Matrix will integrate NVIDIA ConnectX-9 SuperNICs, NVIDIA BlueField-4 DPUs, and NVIDIA Spectrum-X Ethernet networking, creating a proven foundation for deploying specialized inference solutions.
The adoption of NVLink Fusion is part of NVIDIA's broader efforts to accelerate the adoption of AI and machine learning across industries. The company's full-stack AI factory platform, which includes the NVIDIA Vera Rubin NVL72, Groq 3 LPX, and Vera CPU rack, is designed to be completely fungible, running every AI workload, model, and model architecture, with the best performance per watt and lowest cost per token.
By opening up its AI factories to specialized XPU architectures, NVIDIA is providing a platform for silicon innovators like d-Matrix to increase performance, accelerate time to market, and reduce the risk of deploying semi-custom AI factories.
This strategic partnership between d-Matrix and NVIDIA marks an exciting development in the AI inference space, and is expected to have a significant impact on the industry's adoption of AI and machine learning solutions.
Related Information:
https://www.digitaleventhorizon.com/articles/d-Matrix-Adopts-NVIDIA-NVLink-Fusion-for-Accelerated-AI-Inference-and-Deployment-deh.shtml
https://blogs.nvidia.com/blog/d-matrix-nvlink-fusion/
Published: Thu Sep 10 13:07:54 2026 by llama3.2 3B Q4_K_M
