The pace of silicon innovation demands infrastructure that doesn’t slow you down. Semiconductor companies, design service providers, and electronic design automation (EDA) tool developers are under pressure to reduce time to market, increase design complexity, and manage global collaboration—without sacrificing simulation throughput, tool licensing efficiency, or data integrity.
As the demands of EDA workloads grow, traditional on-premises compute, storage, and networking become bottlenecks, prompting semiconductor manufacturers and design teams to seek scalable cloud infrastructure that meets their capacity, performance, and flexibility requirements. Increasingly, they’re turning to Google Cloud—not just for its raw compute and storage capabilities, but because of its ongoing investment in developer-friendly, high-performance infrastructure built for agility, integration, and global scale.
The transition to 5nm, 3nm, and even sub-2nm process technologies has significantly increased design complexity, pushing the limits of traditional infrastructure. More design rules, more simulation cycles, and higher fidelity verification mean that EDA workloads are growing not just in scale, but in intensity as well.
Among the offerings driving this shift is Google Cloud NetApp Volumes, a fully managed, cloud-native file service built on NetApp® ONTAP® and jointly engineered by Google Cloud and NetApp to combine industry-leading file system technology with Google’s scalable, developer-friendly infrastructure. NetApp Volumes is POSIX-compliant and purpose-built for high-throughput, high-scale engineering workloads like EDA. It delivers the scale, speed, and operational simplicity required to support next-generation chip design across front-end logic development and back-end verification.
Infrastructure requirements of EDA workloads
EDA workflows push infrastructure to its limits, demanding extreme compute parallelism, sustained bandwidth, and file systems that can keep pace with complex I/O behavior. Front-end design phases often spawn thousands of short-duration jobs that hammer storage with metadata-heavy operations. Back-end stages like physical design and timing analysis require high-throughput sequential reads and writes across terabyte-scale datasets. These workloads must be completed under tight timelines and with predictable performance to avoid disrupting licensing efficiency or jeopardizing tapeout milestones.
Beyond raw performance, EDA infrastructure must also accommodate a wide range of toolchain behaviors, from stat-heavy small-file operations to multi-threaded, long-running simulations that tax both compute and storage layers. It must offer flexible access models, support shared namespaces across thousands of cores, and allow seamless scaling, both up and out.
Architecture-aware performance for every EDA phase
Silicon design workflows vary dramatically in their I/O profiles. During logic design and IP integration, jobs generate and consume millions of small files, requiring ultra-low-latency access and fast metadata resolution. As projects move into synthesis, place and route, and final signoff, the storage burden shifts to large sequential file access and sustained bandwidth.
NetApp Volumes is designed to manage both ends of this spectrum. The introduction of large-capacity volumes on the Premium and Extreme service levels supports larger datasets, up to 1PiB per volume. Each volume can be mounted by using six parallel endpoints, providing balanced client-side access to sustain high-concurrency simulation pipelines. NetApp Snapshot™ and clones further enhance efficiency, enabling fast checkpointing and rollback without interrupting compute jobs.
The following graph illustrates the output from performance testing of an industry-standard benchmark for EDA. For this workload, the benchmark represents the I/O blend typical of a company running both front- and back-end EDA workloads in parallel.
As the following two graphs show, a single large volume can support up to 21GiB/s of client-side throughput as well as the associated IOPS and latency.
Built for parallel scale
At the heart of NetApp Volumes’ large-capacity volumes is NetApp ONTAP FlexGroup technology, which presents a single, scale-out namespace with automatic load balancing across constituent storage components. This capability enables organizations to run EDA workflows at massive scale without needing to shard data or rearchitect pipelines. Verification teams can spin up tens of thousands of cores and mount the same shared volume from Google Cloud VMware Engine and Compute Engine virtual machines, Google Kubernetes Engine, or preemptible compute instances.
Linear performance scaling ensures that as your design workloads grow, your storage can keep pace—reliably and predictably. Each large volume delivers the throughput needed for high-speed regression testing, physical signoff, and iterative simulation cycles, all without the overhead of traditional storage management.
Recently updated 100% sequential READ throughput maximum for streaming workflows on large capacity volumes has increased from 12.5GiBps to 30GiBps of throughput performance.
Consolidated, cost-optimized infrastructure
NetApp Volumes doesn’t just solve for performance—it simplifies storage operations. With built-in auto-tiering, infrequently accessed files are automatically moved to lower-cost tiers, reducing spending by up to 80% without manual intervention. The fully managed control plane integrates with familiar tools like the Google Cloud Console, Terraform, and RESTful APIs, offering automation and visibility at scale.
Rather than managing a patchwork of file servers or dealing with fragmented volumes, design teams can consolidate storage onto fewer, larger volumes, minimizing complexity and freeing engineering time to focus on what matters: faster design closure.
Designed for the cloud-native silicon era
Google Cloud NetApp Volumes is part of a broader ecosystem optimized for EDA and high-performance computing (HPC) workloads. With elastic compute options such as C2D and HPC-optimized virtual machines, integrated Kubernetes support via Google Kubernetes Engine, and powerful analytics and machine learning capabilities through Vertex AI and BigQuery, engineering teams can build high-performance design environments that are both scalable and agile.
NetApp Volumes also supports Assured Workloads, customer-managed encryption keys (CMEK), identity and access management (IAM), LDAP, and Virtual Private Cloud Service Controls (VPC-SC) for secure, compliant operations, which are crucial in regulated industries and global teams.
Although adjacent services like AI and Kubernetes add significant value, the foundation remains NetApp Volumes: a file service purpose-built to deliver petabyte-scale I/O with operational simplicity.
What’s new for EDA
Recent updates make NetApp Volumes more capable than ever.
Large-capacity volumes now support up to 1PiB per volume and 10PiB per storage pool, dramatically reducing the need to manage multiple datasets or manually provision storage across projects. Performance scales linearly up to 30GiB/s of pure sequential read with the Premium and Extreme service levels, ensuring that your workloads aren’t gated by infrastructure limits. Cross-region replication adds built-in disaster recovery and supports globally distributed collaboration. Premium and Extreme service levels are now available in 14 regions globally, giving customers broader access to NetApp Volumes’ performance profile.
Technical architecture overview
NetApp Volumes integrates seamlessly with Google Cloud’s scalable compute and data services to support EDA pipelines from simulation to analytics. The following figure illustrates a common high-performance EDA reference architecture that leverages NetApp Volumes.
In this environment:
- NetApp Volumes (NFS mounts) stores tools, libraries, RCS, and workspace data across scalable shared volumes.
- Autoscaling managed instance groups of regular and preemptible Google Compute Engine VMs run simulation and verification workloads.
- Remote desktops and license servers provide developer and toolchain access.
- Object storage manages I/O staging, while logging, monitoring, BigQuery, and Vertex AI power downstream analytics and insight.
Key technical specifications include:
- Throughput. Up to 21GiB/s per volume (Extreme service level) client side for EDA workloads
- Volume size. Up to 1PiB per volume, 10PiB per pool
- Protocols. NFSv3, NFSv4.1, SMB
- Mounting. Up to 6 parallel client connections per large volume
- Snapshot and clones. Instant, user recoverable, and automation friendly.
This architecture provides the scale, performance, and simplicity required for modern chip design pipelines, whether the priority is bursty front-end workloads or sustained back-end throughput.
Next steps
If you're evaluating options to modernize your silicon design infrastructure, start by exploring how Google Cloud NetApp Volumes supports the scale and complexity of today’s EDA environments. Whether you're taping out a 5nm chip, validating IP blocks, or running parallelized machine-learning-driven verification cycles, NetApp Volumes is purpose-built to move your workflows forward.
Want to go deeper?
- Read the NetApp Technical Blog on Google Cloud NetApp Volumes for EDA for more architectural detail, benchmarks, and workload-specific guidance.
- Connect with your Google Cloud or NetApp team to discuss your roadmap and see how NetApp Volumes fits into your broader EDA transformation. It’s a unified file service that delivers the IOPS, throughput, and manageability that EDA teams demand, without compromise.
Get started with Google Cloud NetApp Volumes today; or connect with our Google Cloud team to explore how it fits your EDA infrastructure strategy.