The next chapter of AI will be written in systems, not silicon
That was a clear message from Nvidia’s GTC conference last week, the formerly academic conference that has now become a mix of technical sessions and world-class trade show. As one of the newer entrants to our Tercera 30, the Tercera team attended the conference to better understand how Nvidia and others are pushing the boundaries of what’s possible with AI, robotics and accelerated computing. Beyond sessions, we also met with analysts, channel and services leaders to gain insights into the partner ecosystem and the implications for IT services firms looking to navigate the evolving landscape of enterprise AI adoption.
Here are five high-level takeaways that stood out to us — and why they matter for service providers of all sizes:
1. AI Isn’t Just a Tool — It’s a Flywheel for Your Business
One of the most compelling themes across multiple sessions was the concept of “agent-powered flywheels” within organizations. These aren’t just AI tools for productivity—they’re autonomous or semi-autonomous agents that learn, act and improve processes over time. When implemented effectively, they can kickstart a compounding cycle of efficiency, insight and innovation.
For IT services firms, this means rethinking how you structure internal operations and customer-facing processes alongside agents. Could you automate onboarding, and let an AI agent handle repetitive security checks or internal documentation updates? Could you leverage agents to smooth out and accelerate the sales to delivery handoff, or identify areas for future work? The internal use of AI isn’t just cost-saving—it becomes a demonstration of what you can do for your clients.
2. From Generative AI to Agentic AI — and Beyond
The next wave of AI is already here. We’re quickly moving from passive generative models to active, agentic AI — systems that can autonomously take action based on context and goals. Nvidia emphasized this shift as a foundational change, with implications ranging from enterprise automation to real-world robotics and digital twins.
For IT services firms, this represents a major expansion of scope. Projects will no longer just focus on integrating LLMs but designing workflows where AI agents can make decisions, interact with environments (both digital and physical), and continuously improve outcomes.
3. AI Adoption is Inevitable — Control the Chaos Before It Controls You
Another major takeaway: your employees are already using AI, whether sanctioned or not. Shadow AI—tools like ChatGPT, Gemini, or Copilot used without IT oversight—is everywhere. Rather than resist, the recommendation from Nvidia’s Chief Security Officer, David Reber, was to lean in and establish centralized AI governance frameworks.
This is a golden opportunity for services firms to help customers architect secure, scalable and compliant AI strategies. Internally, invest in your own AI policy and infrastructure, then use it as a blueprint for client engagements.
4. The Big Opportunities Aren’t Just for the Big Players
While the Global Systems Integrators are clearly well-positioned to lead in this space, Nvidia’s ecosystem still leaves room for nimble, specialized players — especially those who bring deep vertical or domain-specific expertise.
However, the ecosystem around Nvidia and other major AI players is still relatively closed and extremely difficult to break into, especially if you don’t bring a brand or customer weight to the table. Smaller firms can still play here, but those looking to break in here will need to be proactive, build their own pipeline and give it time. Embedding AI into your core offerings — rather than spinning up standalone AI practices — will be key. Co-innovation, proof-of-concept development, and managed AI services are all emerging as viable paths forward.
5. Nvidia is Evolving—And So Should You
Perhaps the most striking signal at GTC was Nvidia’s own transformation. While GPUs are still foundational, the company is clearly leaning into software, robotics, and full-stack AI platforms like Nvidia Omniverse and CUDA-X. Part of this is due to supply chain constraints (they can only make so many chips), but it’s also a recognition that the next chapter of AI will be written in systems, not silicon.
It will be interesting to see how the tight partnerships between Nvidia and the Cloud Service Providers (CSPs) evolve over the next few years, as Nvidia moves up the stack and the CSPs look to develop their own chips.
The good news is that Nvidia understands that their evolution and the future of AI will require a robust ecosystem of software partners, systems integrators, and service providers. They were vocal that they can’t do this alone. For IT services firms, this is an open invitation to innovate, collaborate, and co-build with one of the world’s most forward-thinking tech companies.
Final Thoughts
AI is no longer a future consideration—it’s a present-day imperative. At GTC, it was clear that Nvidia is not just building the infrastructure of AI, but also rallying an ecosystem to shape its application across industries. For IT services companies that can bring real expertise and differentiated offerings to the table, it could be a good time to step into that ecosystem but don’t expect it to be easy. However, the opportunity s enormous.
