Reshaping Workflows with Dell Pro Precision and NVIDIA RTX PRO GPUs · 2026-08-20 · 32 min
Is slow model performance stopping your AI projects from scaling? In this episode of Reshaping Workflows with Dell Pro Precision and NVIDIA RTX GPUs , Logan Lawler sits down with Stefano Ermon, academic and Inception CEO, to uncover how diffusion-based LLMs are breaking speed records, generating tokens up to 10x faster than traditional models. Stefano Ermon walks us through the evolution from academic research at Stanford to launching the Mercury 2 model, now powering high-speed, latency-sensitive enterprise applications. From the science of parallel generative processes to enterprise-grade deployment on NVIDIA hardware, learn how Inception’s models are changing the rules for AI inference, cost, and scalability. You’ll also hear real-world applications across voice agents, code generation, and search, plus candid discussion about the future of AI model specialization as hardware rapidly advances. Watch now and see how AI is getting infinitely faster! You can also watch this and all previous episodes here .
