Healthcare organizations are facing a delicate balance between tight budgets and the need for operational transformation. While finances remain precarious, with year-to-date operating margin increases hovering below 0.5%, the industry is increasingly turning to artificial intelligence. Two-thirds of healthcare organizations are actively using AI, and nearly half are assessing or using agentic AI.
The AI Deployment Dilemma
The question of where and how to best deploy AI is a pressing one. Tony Nunes, senior manager for strategic business development in healthcare and life sciences at AMD, highlights the challenge of balancing AI workloads on the right infrastructure while managing costs. With cloud management costs remaining high, this strategy is key but can overlook the importance of aligning and redesigning workflows to accommodate AI.
Nunes emphasizes that implementing AI without considering workflow changes can lead to resistance. He notes, “Whether you build something in-house or buy off the shelf, you’re investing in something that changes how people, processes, and technology work together. If you’re not prepared, you’ll meet resistance.”
Motivations for AI Adoption
Three key factors drive AI adoption in healthcare: automation, edge computing, and the fear of missing out. Automation can significantly reduce costs by streamlining operational workloads, patient interactions, and clinical processes. Edge computing, which shifts AI capabilities to client devices, reduces latency and enables real-time data analysis.
The fear of missing out also plays a significant role. Nunes observes, “Corporate buyers can be just as emotional as consumers. When there’s a chorus of voices talking about generative AI, organizations fear being left behind.” This urgency often leads to retrofitting AI tools into existing workflows, which can cause inefficiencies.
When these factors drive decision-making, organizations may overlook the need for workflow realignment. Nunes points to MIT research showing that many organizations struggle to see a return on investment for generative AI. He explains, “Healthcare workflows have specific staffing requirements. If AI doesn’t align with these, new workflows emerge, requiring a different mix of people that no one had planned for.”
The persistence of legacy infrastructure further complicates AI adoption. Nunes notes that this makes it difficult to optimize performance, hindering cost reductions. For example, the evolution of electronic health records has introduced AI orchestration layers to overcome data insight barriers. Nunes stresses, “When introducing new infrastructure, alignment is key. Without it, costs increase.”
Faster refresh cycles for software systems, often driven by AI functionality, add another layer of complexity. Organizations face three choices: refresh infrastructure, add another platform, or wait. Given the costs, many opt to wait, but this approach may not be sustainable as long-term cost reduction becomes imperative.
Nunes suggests that processors should be the first component of modernization, serving as a strong foundation. Flexible and secure infrastructure is ideal, allowing seamless expansion and cost minimization.
Infrastructure built on open standards, as opposed to proprietary systems, offers significant advantages. AMD has enabled the migration of at least 90% of workloads to open stacks.
The Role of Processors in Modernization
Processors are seen as the foundational element in modernizing healthcare infrastructure. This approach ensures that organizations can achieve the same performance with fewer core servers, freeing up valuable data center space.
Open Standards vs. Proprietary Systems
The shift towards open standards in infrastructure is gaining momentum. Nunes draws a parallel to the transition from monolithic mainframes to client-server systems, which occurred over 30 years ago.
Open standards not only close the performance gap but also provide access to a broader partner ecosystem. This expanded network supports innovation and offers consumers more choices, ultimately driving down costs. Nunes concludes, “We’re betting that open standards will provide more opportunity for development and more choice on the part of consumers—and both will lead to reduction of costs.”
