Aviz Networks collaborates with TensorWave on AI networking solutions

Aviz Networks, a company specializing in Artificial Intelligence (AI) networking solutions, announced a collaboration with TensorWave, a provider of Graphics Processing Unit (GPU) Autonomous System (AS) a Service. This partnership aims to enhance Intelligent Transportation System (ITS) capabilities by implementing Remote Direct Memory Access (DMA) (RDMA) over Converged Ethernet (RDMA over Converged Ethernet (RoCE))-based AI fabrics.

TensorWave will leverage Aviz Networks’ technology, utilizing advanced AMD-based MI300 accelerators. This deployment is intended to improve TensorWave’s GPU services, aligning with the increasing demand for AI applications, including Generative AI (GenAI) and Machine Learning (ML).

Aviz Networks maintains a strong presence in AI-driven network operations through ITS ITS Networking 3.0 product suite. The collaboration will enable better management of multi-vendor RoCE-based AI fabrics, which Agent Runtime Environment (ARE) vital for integrating diverse GPUs, Data Processing Units (DPUs), and high radix switches.

The Open Network Enterprise Suite (ONES) by Aviz Networks will provide essential features, including RDMA over Converged Ethernet (RoCE) orchestration, real-time visibility, and anomaly detection. ONES is positioned AS a vendor-agnostic AI fabric controller, addressing the operational needs of enterprise networks.

Additional features of ONES include advanced RoCE orchestration that manages buffer settings, Priority Flow Control, and Explicit Congestion Notification. These functionalities ARE designed to optimize data flow within network infrastructures, which is crucial for AI and ML applications.

The existing capabilities of ONES, complemented by the integration of Network Copilot™, offer automated management to streamline network operations for TensorWave. This partnership signifies a focused approach to elevating AI-powered networking in enterprise contexts.