Navigating the semiconductor revolution

In today's fast-paced AI, automotive, and consumer electronics market, accelerating electronic design automation (EDA) flows is essential for businesses. Chip design teams are constantly updating and modernizing their design flows just to keep up with the growing complexity and size of 3-nanometer (3nm) and now 2nm design processes. These challenges push engineers, EDA tools, and IT infrastructure to the limits.

As digital designs get larger and process technologies get more complex, the demand for IT infrastructure increases and becomes more challenging. The industry has already experienced 4x-6x growth in CPU cores and 4x storage increases per project moving from 5nm to 3nm. According to leading design teams, moving to 2nm will require another 4x increase in CPU cores and storage capacity per project. For example, if a 5nm design required 10k cores and .25PB of data, a 3nm project required 40k cores and 1PB of data. A 2nm project would be projected to require 120k cores and 4PB of data per project.

This explosive growth has put strains on data centers as they reach their power and rack space limits. If you add in the need for new AI GPU infrastructure to power agentic AI–driven design flows, existing data centers are hitting their physical limits. Agentic AI–driven design flows are critical for closing the engineering productivity gap caused by the lack of college grads entering the semiconductor market.

These challenges have led to exploding chip development costs, aka nonrecurring engineering (NRE) costs. NRE cost is the total project cost, which includes people (~70%), design IP and EDA licenses (20%), and IT infrastructure (10%). Reducing NRE is critical because profitability typically requires a 10x return on NRE.

The cloud, either public (AWS, Microsoft Azure, Google Cloud) or private, offers the ability to quickly access nearly instant elastic compute capacity. EDA workloads tend to be bursty in nature, so being able to quickly scale up capacity to meet workload demand is ideal. Cloud is designed to enable quick scale-up and scale-down capacity on demand. On-premises compute resources are fixed, leading to periods of both high and low demand that affect overall data center utilization, but more importantly affect engineering productivity when compute availability exceeds compute demand.

Cloud provides more CPU choice in terms of processor types, sizes, and cost. Cloud providers typically have the latest CPUs from Intel, AMD, and ARM before most companies. Moreover, cloud provides a broader range of available compute configurations, performance needs to dual protocol support, dynamic service levels, and volume resizing.

Cloud is also uniquely ready to quickly enable adoption of GPU-heavy agentic AI workloads today. As new EDA workflows become available from Synopsys and Cadence, design teams will need access to the hard-to-come-by GPUs.

Cloud, either in a hybrid cloud or all-in-one-cloud model, will be required to meet the rapidly growing demand for compute capacity and GPUs for agentic AI–driven productivity, improved engineering productivity, lower project NRE, and project profitability. What often gets lost is improved productivity, which also leads to higher quality, higher levels of innovation, and more predictable business outcomes.

Based on the data collected over several years by the NetApp cloud team, organizations often migrate high-performance computing (HPC) workloads to the cloud to scale compute on demand, to reduce the complexity of on-premises infrastructure, to improve performance, and to manage costs flexibly.

According to 2025 Design Automation Conference (DAC) data, agentic AI will play a crucial role in addressing chip design complexity and the shortage of skilled engineers by providing smarter automation. Customers are seeking robust hybrid cloud solutions to prepare for the demands of agentic AI.

For nearly 30 years, the semiconductor industry has relied on NetApp to provide fast, scalable, secure, and efficient data management for its EDA workflows.  NetApp^® ONTAP^® software is also the semiconductor industry’s choice in the cloud. ONTAP is available as native managed cloud services—Amazon FSx for NetApp ONTAP, Azure NetApp Files, and Google Cloud NetApp Volumes—as well as in a self-managed version called NetApp Cloud Volumes ONTAP, available in all three clouds.

The following figures illustrate NetApp’s intelligent data infrastructure for EDA in the cloud and an architecture example.

ONTAP in the cloud provides the same level of performance, security, connectivity, and scalability as the industry has relied on in its on-premises data centers. The EDA data management challenges can be seen from two different lenses: that of the engineering teams and that of the IT/storage teams. Each has unique, complementary requirements.

From the engineering lens, data must reside on an enterprise-grade, reliable, fast, scalable, simple, and consistent system that enables engineers to run their EDA workloads reliably and consistently. Data must be available where they need it and when they need it, whether it’s in a center supporting the United States, India, or Germany, or in a specific cloud region. Waiting for data slows development and innovation.

From the IT/storage lens, data must provide fast, consistent performance while also being secure, protected, reliable, and available—and simple to manage. With data measured in tens to hundreds of petabytes, data manageability at scale is a hard requirement. Because most semiconductor design teams and design data follow the sun in many different countries, data must also be manageable and secure wherever the data lives, whether that’s in a data center or in the cloud.

Leveraging cloud data management solutions based on intelligent data infrastructure is key to achieving these goals. These solutions offer better cost optimization, because you can maximize business value and minimize unnecessary expenses. On-premises clusters may be capacity-constrained or require large capital expenditure (capex) investments, and they might not be able to keep up with the latest compute and storage hardware cost-effectively. Cloud provides better performance efficiency, data consistency, and business continuity. Cloud also supports better parallel I/O operations, which is critical for workloads that require concurrent access to large datasets.

• Hybrid cloud and cloud readiness: EDA workflows generate and process huge volumes of data that need to move seamlessly between on-premises and cloud environments—and bottlenecks or delays can kill productivity. ONTAP and NetApp FlexCache® features solve this problem by offering unified data management and the ability to cache active datasets wherever the compute is. FlexCache can securely, continuously, and instantly keep all design data synchronized, both on premises and in the cloud. The key features from both products enable engineers to get instant and secure access to the data they need, whether they’re running workloads on premises, in the cloud, or both. No other vendor offers this level of hybrid cloud readiness, flexibility, and security. Engineers often refer to this capability as “single namespace,” meaning that the data is always securely accessible anywhere.
• Scalability, reliability, and flexibility: NetApp cloud storage solutions provide unparalleled scalability. Whether your EDA workload is small or large, NetApp enables you to scale your storage needs easily. The legendary reliability of ONTAP and the always-in-sync architecture of FlexCache mean that your data is available whenever and wherever it’s needed. This flexibility is crucial for businesses that experience fluctuating data volumes and need to adapt quickly without compromising performance. For simulation or analytics workloads that exceed tens or hundreds of terabytes, NetApp supports large volume sizes with performance to match, enabling petabyte-scale storage for massive datasets. This eliminates the need to manage multiple mount points, reducing complexity and performance overhead.
• High performance and low latency: EDA workloads demand high performance and low latency to process and analyze data efficiently. NetApp cloud storage, based on intelligent data infrastructure solutions, is designed to deliver lightning-fast data access and processing speeds—enabling your EDA applications to run smoothly, providing real-time insights, and enhancing decision-making capabilities. NetApp technology in the cloud delivers submillisecond latency and high throughput, making it ideal for performance-intensive workloads such as HPC, analytics, and databases. Its architecture supports parallel I/O operations, which is critical for workloads that require concurrent access to large datasets.
• AI-ready: Agentic AI is set to reshape EDA and semiconductor design, closing the productivity gap and allowing new levels of automation and innovation. But none of this is possible without the intelligent data infrastructure. With industry-leading data mobility, security, and availability features, NetApp intelligent data infrastructure in the cloud is ready for agentic AI EDA workloads. Whether organizations are bursting workloads to the cloud or collaborating across continents, their AI agents and their engineers can count on NetApp.
• Robust protection and security: In a world where chip tape-outs are measured in millions of dollars per day, downtime is not an option. With pointer-based NetApp Snapshot™ copies in the cloud, you can reduce overhead and facilitate quick partial/whole-volume restores. And because EDA data is among the most sensitive in the world, you need robust security when you move that data between environments. NetApp’s intelligent infrastructure in the cloud provides enterprise-grade security and comprehensive data protection features, including encryption, ransomware protection, access controls, and regular security updates. This security framework ensures that the organization’s sensitive data is safeguarded against breaches, providing peace of mind as you automate your electronic data processes.
• Seamless integration with cloud services: NetApp's intelligent data infrastructure and cloud storage solutions integrate seamlessly with major cloud service providers such as AWS, Azure, and Google Cloud. This interoperability allows you to leverage the best of both worlds—NetApp's advanced storage capabilities and the extensive services offered by cloud providers—which simplifies data management and enhances the overall efficiency of your EDA workloads.
• Cost efficiency: NetApp's cloud storage solutions offer cost-effective options that align with your budget. First, these solutions enable predictable cost modeling and encourage rightsizing to avoid overprovisioning, and reserved capacity pricing for long-term projects provides significant discounts. Next, Snapshot copies are space-efficient and instantaneous, enabling frequent recovery points without duplicating data or incurring significant storage overhead. In addition, the data tiering feature lets organizations keep EDA active simulation data on high-performance volumes, while keeping older results on cold or object storage, thus saving costs. And you can migrate older or less frequently accessed simulation results to archive tiers, which reduces the overall storage costs but preserves access to historical data when needed.
• Enhanced data management and analytics: NetApp provides advanced data management tools that simplify the organization, retrieval, and analysis of large datasets. These tools are particularly beneficial for EDA workloads, in which quick access to accurate data is essential. NetApp analytics capabilities further enhance your ability to derive actionable insights from your data, driving better business outcomes.
• Global reach and availability: NetApp offers 118 regions of availability in over 30 countries across 3 hyperscalers. With proven high availability, disaster recovery, cloning, replication (NetApp SnapMirror®), and caching (FlexCache) features, NetApp makes sure that you can access your EDA data whenever it’s needed. This global presence means that your data is always accessible, no matter where you are, and it minimizes the risk of downtime, which is critical for continuous data automation processes.

NetApp intelligent data infrastructure in the cloud supports EDA end-to-end engineering workflows. NetApp cloud solutions offer a comprehensive, reliable, and efficient platform for managing EDA workloads. NetApp is the only cloud storage provider that has first-party managed solutions with all three hyperscalers. With its scalability, high performance, robust security, seamless integration, cost efficiency, advanced data management, and global reach, NetApp empowers businesses to optimize their EDA processes and operate more efficiently. Choosing NetApp for an organization’s EDA workloads means choosing a partner committed to innovation, reliability, and excellence in data management.

Contributor: Special thanks to my colleague Michael Johnson for his help with this blog. Michael is a NetApp veteran and former semiconductor engineer, responsible for delivering industry-defining solutions in EDA, automotive, aerospace, and enterprise software.