Introducing Azure NV V710 v5: Real-Time AI/ML Inferencing & Visualization

As industries increasingly rely on high-performance computing and AI for real-time inference, remote operations, and advanced visualization, Azure’s portfolio of virtual machines (VMs) continues to evolve to meet these needs. Today we are excited to introduce Azure NV V710 v5, our latest VM tuned for small to medium-sized AI/ML inference workloads, virtual desktop infrastructure (VDI), visualization, and cloud gaming workloads.

AMD’s latest Radeon™ PRO V710 GPU and 4^Day AMD EPYC™ (formerly “Genoa”) generation high-frequency CPUs deliver high compute performance and flexible GPU partitioning to address a wide range of industry requirements.

Why choose Azure NV V710 v5?

NV V710 v5 brings a new level of flexibility and performance to the cloud, purpose-built for small to medium-sized real-time AI/ML inference workloads and graphics-intensive applications.

Key features of NV V710 v5

Real-time inference (RTI) and AI inference:

NV V710 v5 is optimized for small and medium-sized AI model inference and real-time machine learning processing, providing the computing power and speed needed for industries that rely on instantaneous data processing. Support for vLLM allows users to perform AI/ML inference more efficiently, delivering near-instantaneous results for workloads such as edge AI applications and intelligent decision systems, all while lowering the total cost of SKU ownership. .

GPU partitioning for flexibility:

A standout feature of the NV V710 v5 is its GPU partitioning feature, which allows customers to allocate portions of the GPU based on workload requirements. This flexibility is ideal for multi-tenant environments, allowing organizations to efficiently support a variety of inference and graphics workloads without requiring a full GPU for each application.

High-performance AMD EPYC CPU:

Powered by the AMD 4th Gen EPYC CPU with a base frequency of 3.9 GHz and a maximum frequency of 4.3 GHz, the NV V710 v5 is optimized for demanding computing tasks that require both high CPU and GPU performance. This makes it ideal for complex simulations, graphics rendering, and real-time inference.

Large GPU memory:

With 28GB of GDDR6 GPU memory, NV V710 v5 can handle large-scale model inference, high-resolution rendering, and complex visual content. High memory capacity allows for smooth processing and loading of large data sets in real time.

Azure integration and high-speed networking:

Integrated with Azure accelerated networking, NV V710 v5 delivers up to 80Gbps of bandwidth, ensuring high performance and low latency for AI inference, VDI applications, and cloud gaming workloads. This high-speed networking capability facilitates seamless data transfer and supports intensive graphics and inference workloads.

real application

One of the key applications for the NV V710 v5 is the automotive industry, where AI-based sensor simulation and inference will play a key role in the development of intelligent edge devices for autonomous vehicles. Platforms such as the Automated Driving Perception Hub (ADPH) provide automotive customers with a virtual environment to evaluate various automotive sensors such as cameras, lidar, and radar.

• Correct inference: NV V710 v5 supports batch inference, providing a reliable environment to evaluate AI model accuracy across a variety of simulations.
• Cross-platform support: Compatibility with ROCm/HIP enables cross-platform inference critical to intelligent edge devices.
• A wider range of applications: In addition to the automotive industry, NV V710 v5 can support a variety of edge AI devices, including security cameras, industrial equipment, and drones.

NV V710 v5 Technical Specifications

AI Inference Opportunities with NV V710 v5

NV V710 v5 provides a versatile platform for real-time AI/ML inference and visualization tasks. Support for vLLM allows companies to run complex AI models efficiently in real time, making it an essential asset for industries focused on AI-driven insights. GPU partitioning allows enterprises to optimize resources across diverse workloads, ensuring a cost-effective approach to cloud-based inference and graphics rendering.

Additional use cases

• VDI and remote workstations: For enterprises deploying virtual desktops, NV V710 v5 provides high-performance computing resources that can be dynamically adjusted based on user requirements. This flexibility is useful in media production, design, and financial services where advanced graphics capabilities are important.
• Cloud Gaming: NV V710 v5 is built to handle cloud gaming with low-latency performance, giving gamers a seamless, high-quality experience comparable to traditional gaming consoles. Powerful architecture supports real-time rendering to deliver premium gaming experiences in the cloud.

Conclusion: The future of AI inference and graphics workloads with Azure NV V710 v5

Azure NV V710 v5 VM is set to transform the landscape of AI inference, real-time visualization, and cloud gaming. It combines a high-performance AMD Genoa CPU, 24GB GPU memory, ROCm 6 support, and vLLM to provide an all-in-one solution for a variety of applications.

NV V710 v5 opens new opportunities for enterprises to run real-time AI/ML model inference in the cloud, efficiently scale graphics workloads, and deliver high-quality user experiences. Tailored to meet the needs of modern, graphics-intensive, AI-driven industries with advanced partitioning and high-speed networking capabilities.

Are you ready to experience the power of NV V710 v5? Sign up for the public preview here.