SAP is finding NeMo: how the NVIDIA partnership unlocks a use-case arms race

When reading the tea leaves, SAP's partnership with NVIDIA - announced at the same time as NVIDIA NIM - introduces what The Innovator's Dilemma called a "disruptive innovation": a technology so game-changing that it provides immediate competitive advantage.

It isn't just about a partnership with the biggest AI company in the world. So ... stop the presses and change the headlines to read the following:

SAP is rolling out the capability to mass customize fine-tuned GenAI for SAP business process use cases across the board, on top of NVIDIA GenAI microservices designed to scale across cloud, on-premise, or even on device.

As Clayton Christensen's book concluded, in a disruption of this magnitude companies that adapt quickly attain market dominance. Laggards, on the other hand are diminished or destroyed.

So ... what do I need to know about the technology, to avoid being diminished or destroyed?

Let's talk about NeMo and NIM

You probably want to checkout the OTHER news release that NVIDIA rolled out. Dozens of AI microservices that are scalable anywhere - even to user's desktops or on-premise datacenters.

Here's a screenshot of the press release:

Contrast this to SAP's announcement, which talked about using the NVIDIA technology to fine-tune the ABAP model.

To the technologists out there who want to avoid being caught off-guard, there are NVIDIA words you'll want to add to your lexicon immediately: NIMS and NeMo. From the 50,000 foot perspective, these are the NVIDIA microservices we'll use to develop, train, and deploy AI use cases anywhere.

Between NeMo and NIM, we have the capability to full stack develop, fine tune, RAG, and deploy LLMs anywhere.

Say hello to NVIDIA NIMs, NeMo.

And the deployment of GenAI is a topic our CIOs have learned a lot about. I hear it everywhere, 16 months after the GenAI wave first splashed across SAP's shores.

Use cases.

I have a co-worker who monitors the birds in his backyard with Bird Buddy. I have other friends who creating synthetic molecules and curing allergies. When you're eating lunch and chatting over schwarma, however, they're both just as enthusastic about the same topic. What else can you use it for?

Now, SAP delivers hundreds of Business AI use cases out of the box, ready for your teams to learn about and deploy. But if you attend any large AI hackathon, any large conference, there are so many team-specific scenarios, where a supply chain assistant, for example, would need to understand federal procurement guidelines from DFARs, cybersecurity guidelines from NIST, or ancient corporate policies, or some other custom bit of information.

So ... how do we deliver bespoke AI use cases to teams today? Where does NVIDIA fit in? Where can we identify a business use cases where we bring our toolbox (AI, analytics, user automation, and user experience) to:

• Redesign TAKT time required for a human to complete a process
• Increase user accuracy
• or to react faster than a human being

How do we scale our business processes to edge compute environments, to a million different use cases, regardless of device or cloud platform?

Finding NeMo: NVIDIA's Technologies Explained

NVIDIA NeMo: NVIDIA NeMo lets you fine-tune or build your own AI models, making it easy to work with different kinds of computer systems underneath. You can start with models that are already out there for everyone to use, or make something completely new that fits exactly what you need. After you've got your model ready, you use NVIDIA NIM to put it into action. NIM helps you deploy your model anywhere you want—on devices, in the cloud, or in containers. This means you can take the AI you've worked on with NeMo and easily get it running where it needs to be, whether that's on a server or right on a local device.

NVIDIA NIM (NVIDIA Inference Microservices): NVIDIA NIM (NVIDIA Inference Microservices) alllow you to seamlessly deploy your trained models across any environment – be it devices, cloud platforms, or containerized solutions.

Leveraging models created with NVIDIA NeMo, NIM enables you to not only deploy these models with ease but also integrate advanced capabilities like retrieval-augmented generation (RAG) through Joule, enriching your applications with deeper insights and more dynamic interactions. With the recent announcement of hundreds of additional microservices, NIM is set to broaden the horizons for deploying AI, offering unprecedented agility and performance optimization across your technology landscape, ensuring that your applications remain responsive and efficient, no matter where they are deployed.

Leveraging the Partnership for GenAI Implementation

Integrating GenAI into Business Processes: With NVIDIA's AI technologies, SAP is enhancing its cloud solutions, such as SAP S/4HANA Cloud, SAP SuccessFactors, and SAP Signavio, with generative AI features. For SAP customers, this translates to a range of possibilities—from automating complex enterprise resource planning tasks with intelligent invoice matching to enhancing HR functionalities and optimizing business processes based on AI-driven insights.

Data Unification and Enhanced Insights with SAP Datasphere: The partnership emphasizes the importance of unified data for AI implementations. SAP Datasphere, supported by NVIDIA's technologies, enables the integration of diverse data sources, offering a comprehensive view for AI models to analyze. This unified data fabric, powered by generative AI, assists companies in adapting faster to market changes and making well-informed decisions.

Development and Deployment Flexibility: For organizations, especially in Regulated Industries, that prioritize data privacy and compliance, the ability to deploy these solutions in your data center is significant. It offers the flexibility to maintain control over sensitive data while benefiting from state-of-the-art AI capabilities.

Empowering Developers with LLMs: The use of Large Language Models (LLMs) for domain-specific scenarios, including programming languages like ABAP, is a game-changer. It means that your development teams can leverage NVIDIA's AI foundry services to fine-tune AI models that understand and automate tasks within your specific business context, significantly reducing the time and effort required for coding and other complex tasks.

SAP and NVIDIA deliver the GenAI microservices you're looking for

Companies are navigating their own journeys of GenAI adoption, but hear me out ... we are at the beginning of a sinus wave cycle of "disruptive innovation" in every field, one that even Clayton Christensen's oft quoted book couldn't predict.

SAP and NVIDIA partnership provides a comprehensive framework to accelerate digital transformation with cutting-edge AI capabilities. By understanding and leveraging NVIDIA's NIM and NeMo technologies within the SAP ecosystem, you can unlock new levels of efficiency, innovation, and customer engagement. Implementing GenAI is not just about adopting new technology; it's about learning to compete in a whole new ball game.

Thanks for reading my article!

Learn more in the press releases about the SAP and NVIDIA partnership, about NVIDIA NIM, and learn more about SAP Business AI use cases at SAP.com.