Microsoft and NVIDIA are taking a major step forward in using AI to accelerate scientific discovery and innovation across industries. By combining their strengths, the two tech giants are making it easier for researchers to tackle complex problems from drug discovery to materials science faster than ever before.

At the heart of this collaboration is agentic AI, a new class of AI systems designed to act with greater autonomy and intelligence. These systems are helping scientists move from idea to breakthrough in record time, whether they’re working in the cloud or on a local machine.

Microsoft Discovery: A New Platform for R&D Teams

Introduced at Microsoft Build, the new Microsoft Discovery platform is designed to help research and development teams reinvent the discovery process. The platform brings together powerful AI tools that can streamline everything from early research to product development.

Whether it’s speeding up simulations, finding the best drug candidates, or identifying new materials, Microsoft Discovery is built to accelerate results and bring ideas to market faster.

Powered by NVIDIA: Smarter Tools for Chemistry and Biotech

To supercharge the platform, Microsoft is integrating NVIDIA’s ALCHEMI and BioNeMo NIM microservices.

• ALCHEMI NIM helps with chemical simulations, making it easier to predict how new materials will behave.
• BioNeMo NIM taps into pre-trained AI workflows tailored for life sciences, offering researchers an efficient way to explore new drug therapies.

One recent example shows the platform's impact: Microsoft researchers used Discovery to identify a promising new coolant for data centers and they did it in under 200 hours, a process that normally takes months or longer.

Scaling AI Power in Azure with NVIDIA GB200 Systems

To support this leap in AI capabilities, Microsoft is rolling out tens of thousands of NVIDIA GB200 NVL72 systems across its Azure data centers.

These high-performance systems feature.

• Up to 72 NVIDIA Blackwell GPUs per rack
• Advanced liquid cooling for energy efficiency
• Massive throughput — delivering 35x more inference performance than the previous generation

It’s a serious upgrade aimed at handling today’s most demanding AI workloads — including generative and agentic AI applications.

Optimizing Performance, Together

During the keynote, Satya Nadella and Jensen Huang highlighted how this deep collaboration goes beyond hardware. Continuous software improvements across NVIDIA architectures on Azure are helping developers get more done, for less.

These optimizations.

• Improve developer productivity
• Lower infrastructure costs
• Deliver better energy efficiency per dollar and per watt

More AI Power for Healthcare and Reasoning

Microsoft and NVIDIA are also enhancing Azure AI Foundry by expanding support for advanced reasoning and healthcare tools.

This includes.

• NVIDIA Llama Nemotron models for multistep reasoning, code generation, and agentic tasks
• BioNeMo NIM services like ProteinMPNN, RFDiffusion, and OpenFold2, designed for biotech and medical applications such as protein folding and genome analysis

These additions are designed to help researchers in healthcare and digital biology move faster with greater accuracy and reliability.

Bringing Generative AI to Windows 11 PCs

On the consumer side, generative AI is coming to the desktop in a big way. NVIDIA RTX AI PCs running Windows 11 are now equipped to handle advanced AI applications right on the device.

The big update here is TensorRT for RTX a redesigned AI engine that,

• Builds inference engines just-in-time on the device
• Uses less memory (with an 8x smaller package size)
• Offers significantly better performance

It’s now supported natively through Windows ML, which is designed to give developers broad hardware compatibility and streamlined performance. The preview is available now, with a full SDK release coming in June via NVIDIA Developer.

Final Thoughts

This partnership between Microsoft and NVIDIA is more than just tech. It's about unlocking the next generation of discoveries. From accelerating science in the lab to powering generative AI on your laptop, these innovations are helping shape a future where progress happens faster, smarter, and more sustainably.