Databricks and NVIDIA: Building for the Agentic Era

2026-06-17 · Source: Databricks · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Cloud Computing & IT Infrastructure, Robotics & Autonomous Systems · Depth: Advanced, medium

Summary

Databricks and NVIDIA are deepening their partnership to accelerate AI workloads across the full lifecycle on the Databricks platform, as highlighted at the Data + AI Summit. This collaboration integrates NVIDIA AI infrastructure into Databricks AI Runtime, Model Serving, and Industry AI solutions. Key developments include AI Runtime's support for NVIDIA Hopper GPUs with NVIDIA Quantum InfiniBand for distributed training, with future readiness for NVIDIA Blackwell architecture, and the introduction of GPUs in Databricks Free Edition. For inference, Model Serving leverages NVIDIA hardware and Triton Inference Server for high-throughput, low-latency performance. The new NVIDIA Vera CPU is designed to power agentic infrastructure, offering up to 3x faster SQL queries and 80% faster agentic performance for latency-sensitive tasks. Additionally, the NVIDIA Agent Toolkit can be deployed on Databricks Apps, and Genie Code provides conversational debugging for GPU workloads. The partnership also extends NVIDIA's domain-specific libraries, such as NVIDIA MONAI and NVIDIA BioNeMo, to Databricks for specialized industry AI applications.

Key takeaway

For AI Engineers building or deploying agentic AI workflows and large-scale models, this expanded Databricks-NVIDIA partnership provides a unified, accelerated platform. You should explore utilizing Databricks AI Runtime with NVIDIA Hopper GPUs for training, and consider the new NVIDIA Vera CPUs for agent orchestration to overcome CPU bottlenecks. Deploy the NVIDIA Agent Toolkit on Databricks Apps for streamlined agent development, and use Genie Code for efficient GPU workload debugging and optimization, ensuring predictable performance and governance for your enterprise AI initiatives.

Key insights

NVIDIA's full-stack AI acceleration, including new Vera CPUs, integrates deeply with Databricks to power enterprise AI and agentic workloads.

Principles

• Agentic workloads benefit from purpose-built CPUs for orchestration.
• Full-stack hardware-software integration optimizes AI lifecycle.
• Governed data platforms are crucial for enterprise AI.

Method

Deploy NVIDIA Agent Toolkit on Databricks Apps for agentic AI workflows, leveraging built-in authentication and governance. Utilize Genie Code for conversational debugging and performance optimization of GPU workloads.

In practice

• Use NVIDIA Hopper GPUs with NVIDIA Quantum InfiniBand for distributed training.
• Leverage Databricks Free Edition for GPU-accelerated AI development.
• Employ NVIDIA Vera CPUs for agent orchestration and data analytics.

Topics

• AI Agents
• GPU Acceleration
• Databricks Platform
• NVIDIA Vera CPU
• Enterprise AI
• AI Infrastructure

Best for: AI Architect, CTO, VP of Engineering/Data, AI Engineer, Machine Learning Engineer, MLOps Engineer

Related on AIssential

• Nvidia introduces Vera Rubin, a seven-chip AI platform with OpenAI, Anthropic and Meta on board — Nvidia's Vera Rubin platform and ecosystem target agentic AI with a comprehensive, integrated hardware and software stack.
• Introducing AI Runtime: Scalable, Serverless NVIDIA GPUs on Databricks for Training and Finetuning — AI Runtime simplifies distributed GPU training by offering serverless, on-demand access and optimized tools within Databricks.
• NVIDIA Vera Rubin POD: Seven Chips, Five Rack-Scale Systems, One AI Supercomputer — NVIDIA's Vera Rubin POD integrates five specialized rack systems into a cohesive AI supercomputer for agentic AI.
• NVIDIA Partners With Microsoft on Unified Stack for Agentic AI Deployment, From Windows Devices to Cloud to Local — NVIDIA and Microsoft are building a full-stack, secure, and scalable ecosystem for agentic AI, from edge devices to cloud.
• NVIDIA Vera CPU Sets a New Standard for Agentic Workloads in AI Factories — The NVIDIA Vera CPU optimizes agentic AI and reinforcement learning by making CPU execution a high-performance, critical path within the AI loop.
• Cadence, Nvidia expand partnership for accelerated engineering AI — Cadence and Nvidia are accelerating engineering design through agentic AI, digital twins, and accelerated computing integration.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Databricks.