In this fourth episode of AI in Action, hosted by TD SYNNEX in conjunction with NVIDIA, we shift our focus from the physical hardware to the powerful software layer that makes modern AI possible. Join host Ben, along with experts Michael Gern and Jake Fielden, as they demystify the NVIDIA enterprise software ecosystem.

The conversation breaks down how organisations can effectively partition and provision their GPU resources—whether for professional CAD designers, office workers, or intensive AI workloads, using technologies like vGPU and Multi-Instance GPU (MIG). Michael also highlights the "best-kept secret" of NVIDIA software licensing and how developers can access the full toolkit for free to start building today.

Key topics covered in this episode include:

• The Transition to NVIDIA Enterprise: Understanding the rebranding of NVIDIA’s curated software suite and the importance of enterprise-grade support for open-source tools.
• Optimized Models & NIMs: How to utilize NVIDIA Inference Microservices (NIMs) to deploy pre-trained, high-quality models for vision and language tasks without starting from scratch.
• The Ecosystem Advantage: How TD SYNNEX's Destination AI program and ISV partners help bridge the skills gap for companies without in-house data scientists.
• Practical Sizing & Testing: Tips for right-sizing your AI environment and how to access TD SYNNEX AI labs or cloud resources to test your models before you buy.

Whether you are a developer looking for the right tools or a business leader trying to understand the ROI of your hardware investment, this episode provides a clear roadmap for the software stack.

In this episode of AI in Action, a TD SYNNEX podcast produced in conjunction with NVIDIA, our panel of experts dives deep into the physical backbone of the artificial intelligence revolution: Hardware.

As the industry moves beyond experimentation, the conversation shifts to where and how AI actually operates. We explore the full spectrum of the AI hardware stack, breaking down the specific requirements for different environments:

• AI at the Edge: Defining "the edge" as on-site inference in real-world conditions. We discuss the NVIDIA Jetson platform and ruggedized systems from vendors like Lenovo, Dell, and Supermicro that are designed to withstand harsh environments while providing low-latency responses for manufacturing and retail.
• The Developer’s Desktop: Why memory (and specifically Unified Memory) is the most critical resource for developers. We introduce Spark as a specialized execution engine for AI development and discuss the transition path to professional workstations like the DGX Station.
• The Modern Data Center: Moving into large-scale inference and training. Our experts explain the "care and feeding" of high-density systems like the HGX and DGX, the necessity of NVLink to eliminate performance bottlenecks, and the massive power and cooling demands of these 10kW clusters.
• AI-Infused Storage: Why data is the fuel of AI and how partners like NetApp are integrating GPU-powered "AI data engines" directly into storage platforms to handle data preparation, classification, and vectorization.

Whether you are a hobbyist starting on a gaming PC or an enterprise building a full-scale "AI Factory," this episode provides a roadmap for right-sizing your hardware to your specific AI workload.

In this continuation of Episode 2, Ben and the team dive into the practical, high-impact ways AI is transforming industries today. The discussion moves beyond the hype to explore tangible solutions, starting with the evolution of computer vision, from identifying suspicious objects at Houston Station to enhancing athletic performance in professional football. The panel also highlights how AI is becoming a vital tool in healthcare, dramatically accelerating the processing of brain scans to save lives.

As the conversation shifts toward the future, the team breaks down the rise of Generative AI and Large Language Models (LLMs). They explore the critical importance of Sovereign AI, explaining how businesses can leverage techniques like RAG (Retrieval Augmented Generation) and fine-tuning to keep their sensitive data behind their own firewalls. Whether you are a small business looking for a $4,000 entry-level system or a large enterprise scaling up, this episode provides a roadmap for navigating the data-driven landscape of 2026.

Key topics include:

• Safety and Care: How pose estimation is being used for elderly care and fall detection.
• The Mechanics of LLMs: Why these models are "predictive" rather than "intelligent" in a human sense.
• Data Quality: Why every AI project is secretly a "data project in a hat and a false moustache".
• Local Execution: The shift toward processing AI locally to ensure privacy and security.