Powering the future of healthcare: Embracing GenAI in data centers
To enable continual transformation in healthcare, advanced GenAI applications need experienced support and properly configured data centers.
Generative artificial intelligence applications have the potential to simplify and enhance the analysis of medical data at every step of the patient journey. These applications present exciting opportunities to improve patient outcomes, speed up care and reduce errors, but they come with massive data requirements and workloads.
To realize the full potential of GenAI, insurance, pharmaceutical, life sciences and healthcare organizations need data center solutions that can efficiently handle their heavy demands.
Can GenAI revolutionize healthcare?
GenAI opens the door to new applications and innovations that could revolutionize the healthcare experience for patients, medical professionals, insurance providers, pharmaceutical companies and life sciences organizations. By enabling advanced analytics and personalized solutions, GenAI can play a transformative role in enhancing service delivery, therapeutic development and overall health outcomes.
Physicians and nurse practitioners might use GenAI to help diagnose patients, draft patient instructions, update electronic health records or consolidate care coordination notes. These capabilities could free up more time to build relationships with patients and help them create more accurate care plans.
For healthcare insurers, GenAI could automate the claims adjudication process, reducing processing times and improving accuracy. GenAI could also help analyze vast datasets to tailor insurance premiums based on individual risk factors or identify future claims and health outcomes based on historical data.
At the same time, life sciences organizations could leverage it to process and analyze biochemical data to identify potential drug candidates much faster than traditional methods. GenAI could also make it easier to analyze medical records to identify and recruit candidates for clinical trials or develop tailored treatments based on individual genetic profiles.
On the other hand, pharmaceutical companies might employ it to predict potential adverse drug reactions or interactions, optimize supply chain logistics, focus research and development activities or manage production operations. Applications that combine GenAI with computer vision and machine learning could automate quality inspections during manufacturing.
Government and public health organizations could use it to analyze data from various sources to detect and track disease outbreaks in near-real time, predict public health emergencies or optimize the distribution of medical resources. GenAI could also help policymakers analyze large data sets to better understand the impact and effect of planned health policies before implementation.
Patients even could use GenAI chatbots to get answers to routine questions about their health and treatment plans.
These emerging applications — and countless other use cases to come — could be the difference-makers that enable healthcare organizations to address their big challenges. For example, they could help organizations streamline care delivery, lower OPEX, improve patient satisfaction, prevent clinician burnout, manage risk, deliver more impactful population health initiatives and gain an innovative edge. Although concerns about privacy, security and data remain, there is high interest in GenAI across the healthcare industry.
Implementation won’t be easy
The embrace of advanced GenAI applications will put immense pressure on healthcare IT staff and data center networks. These applications must be integrated and maintained by networking teams that are already busy with day-to-day operations and management tasks.
And while GenAI securely automates access to relevant data lakes, its data volumes and workloads can strain existing data center networks. These networks tend to run on proprietary closed operating systems that aren’t easy to customize, integrate or automate for challenging applications.
So, what will it take for organizations to successfully integrate massive data lakes and large language models and get the most from GenAI?
Data centers capable of GenAI
Healthcare organizations can ease the pressures exerted by GenAI applications and unlock their full potential by building data centers that can scale easily and adapt to new and bigger workloads in a much more agile way. This means adopting next-generation solutions that will empower their teams to design, deploy, adapt, operate and automate data center fabrics at a scale to meet the unique requirements of GenAI.
The foundation of a next-generation data center fabric solution is a fully open Linux-based network operating system (NOS) designed for agile, flexible operations. To support advanced GenAI applications, the NOS must be able to ramp up data center network performance, scalability and efficiency with the following features.
- • A cloud-native design that promotes programmability, flexibility and resiliance.
- • An unmodified Linux kernel that can be used to create a suite of network applications.
- • A completely open, model-driven design that gives each application its own data structure.
- • A publisher/subscriber architecture that provides scalable, reliable and secure communications for state sharing and real-time streaming of telemetry data.
- • A state-efficient design that uses microservices to support resilient networking and hitless application upgrades.
- • A programmable management layer that uses a flexible command line interface (CLI) alongside versatile application programming interfaces (APIs) that support JavaScript Object Notation Remote Procedure Call (JSON-RPC) and gNMI to facilitate robust automation, supporting fully customized operations and enhancing DevOps and NetOps agility and efficiency.
- • High-performance IP stacks and network functions that ensure interoperability and security.
- • An open and scalable telemetry interface that doesn’t require translation layers.
Organizations with expectations about GenAI also need tools that can clear the path to more fully automated data center fabric operations. For example, a NOS that supports declarative, intent-based automation can increase efficiency across the data center fabric lifecycle and make it easier to prepare for GenAI.
With an automation toolkit based on open frameworks and microservices, networking teams will be able to meet the scalability and performance needs of different GenAI applications and reduce the operational complexity that comes with them.
Networking teams will be wary of the risks involved in bringing new GenAI-related automations, integrations and workloads into the data center fabric. A solution that provides a data center fabric emulation environment such as a digital twin will reduce these risks by providing a safe virtual space where they can see the impact of changes for GenAI applications before flipping the switch to implement them in the real world.
High-performance data center hardware is a must for healthcare organizations seeking to get the most from GenAI. Merchant silicon-based switching platforms that come in different form factors and share a common hardware and software design will provide the greatest flexibility to address the requirements of GenAI applications.
These platforms should offer high-throughput interfaces that can meet varying speed and scalability demands, as well as Layer 2 switching and Layer 3 routing, telemetry, security, quality of service and model-driven management capabilities.
Reaping GenAI-ready benefits
A data center fabric built with a modern, fully open NOS, flexible automation tools and scalable hardware can empower any organization to realize the transformative potential of GenAI applications.
Equipped with a high-performance, scalable and efficient next-generation data center fabric, a healthcare organization can create a flexible foundation for implementing and customizing network tools that support GenAI applications, as well as reduce the pressure that GenAI puts on networking teams and infrastructure by simplifying and automating every phase of data center fabric operations.
Additionally, they can rapidly and safely develop, test, validate and implement new GenAI-related integrations, configurations and automations to further enhance healthcare — for patients and providers.
Dave McClain is a network solutions architect at Nokia.