IBM Cloud Code Engine introduces Serverless Fleets with GPUs
Published 03 October 2025
?ts=1751827922792&dpr=off "Modern office with collaborative workspace")
To address the unique challenges of running large-scale and compute-intensive workloads on a fully managed, serverless platform, IBM Cloud Code Engine has introduced Serverless Fleets.
Enterprise-grade AI, including generative AI, requires a compute- and data-intensive infrastructure. Serverless technology, in its current form, sometimes falls short of supporting compute intensive workloads.
As technology tackles more complex problems from machine learning to large-scale simulations, systems need to perform thousands, maybe even millions, of tasks often in parallel. Serverless Fleets can help simplify how developers execute these parallel workloads by providing fully managed compute resources that can scale elastically. Traditional workflows for compute-intensive workloads often require upfront sizing and investment in infrastructure and complex resource management.
Serverless Fleets is designed to help reduce this operational complexity by abstracting the underlying infrastructure and providing a simple way to run a large number of tasks in parallel. Serverless Fleets can be game-changing for enterprises across industries including the financial services and life science industries.
Serverless Fleets allows data scientists and developers to submit a large number of batch jobs to a single endpoint. IBM Cloud Code Engine then automatically provisions the necessary compute resources—including virtual machines and serverless GPUs—to run as many tasks simultaneously as possible. This automates resource management that is often a manual process for large-scale calculations.
Think of your current IT environment like a hotel chain. As the owner of a chain, you may have bought several large buildings with thousands of rooms that each have an operating cost and individualized management. Instead, imagine if you could automatically create a room when someone wants to stay at the hotel and then the room is destroyed once the guest checks out. This is exactly how Serverless Fleets operates: provisioning infrastructure for large workloads using GPUs as needed and removing it once it is no longer necessary, and charging you only when the technology is used.
Serverless Fleets builds on the existing value of IBM Cloud Code Engine for compute-intensive workloads such as large-scale simulations, data processing, AI, and machine learning:
- Unified management: Enterprises can manage web applications, batch jobs and serverless functions within a single, unified environment. This central point of control helps to simplify the architecture and management of complex applications that combine different types of workloads.
- Cost optimization: The “pay-as-you-go” model is highly cost-effective for unpredictable or intermittent workloads as Code Engine will provision resources only when jobs are running.
- Security: IBM Cloud Code Engine inherits the security capabilities of IBM Cloud, including data encryption in transit and at rest. This approach is key for managing high amounts of data, especially workloads associated with AI.
Enterprises in highly regulated industries, such as financial services, have a unique set of challenges that must be understood for true transformation to begin. These organizations are tasked with managing sensitive customer data, maintaining resilience and meeting changing regulatory requirements. Running compute-intensive workloads should not be an additional roadblock to innovation.
For example, insurance companies need vast amounts of compute power to run underwriting calculations, build pricing models, report to regulators, and adjust risk in real-time. Serverless Fleets supports these workloads, designed to make it easier and cost-effective for companies in the financial industry to run simulations, models and assessments.
“By embracing the power of Serverless Fleets and IBM Cloud Code Engine, MavenBlue is able stay ahead in our mission of delivering new insights, speed, and analytical agility for the insurance industry” said Paul Hagg, CTO of MavenBlue. “We have a robust set of next generation platforms that provide insights, forward-looking solvency analysis, and pricing models. Serverless Fleets help us to unlock the compute power we need in a flexible and cost-effective manner.”
The life sciences industry is similarly charged with upgrading system performance to remain competitive. Compute power for research and development is key, especially as pharmaceutical companies race to bring new therapies to market, streamline clinical trials and leverage data-driven insights. This requires them to have interoperable systems that prioritize security and reliability. Serverless Fleets can help life science organizations efficiently adapt to the ever-changing market by provisioning the infrastructure necessary without IT teams needing to do so.
“To fulfill Bayer’s mission ‘Health for all, Hunger for none’, we rely on innovative and high-performance solutions in our research and development. With Serverless Fleets, we’ve found a highly scalable and stable platform that enables us to execute even the most demanding compute-intensive tasks reliably—and at very fair conditions. At Bayer, we were particularly impressed by the ease of scaling and the outstanding support provided by IBM,” said Dr. Georg Mogk, Bayer AG, Prinzipal Expert Applied Mathematics.
The reality of today’s competitive landscape is that enterprises across industries need to deliver services quickly and conveniently, while prioritizing security, resilience and cost savings.
Serverless Fleets can help developers and data scientists manage their most demanding workloads to help deliver these services, to give them back time to focus on coding, rather than infrastructure management. As IBM strives to make advance computing more accessible and efficient, Serverless Fleets is a step in the right direction.