Electronic Design Automation
Electronic design environments are often very demanding and complex. They require close collaboration between engineers, designers, and IT staff to successfully complete a project. The complexity comes from application tools and libraries, distributed and mixed network-attached storage (NAS) solutions, and high-end computing clusters and workstations for processing. Management and delivery of data alone is one challenge that continues to strain the design process. Data sets are growing extremely large and are stored in various places throughout the organization. NAS has provided the flexibility and data protection needed to support the operation, while compute clusters have sped the delivery cycle. However, NAS performance has not kept pace with compute capabilities. This is creating an I/O bottleneck which IT is trying very hard to solve.
Every step of the design process requires the use of data from previous stages. Data is constantly read and re-read by engineers and applications. Eliminating NAS I/O bottlenecks in order to provide timely processing, simulations and verifications is a challenge unto itself and can save not only time, but money as well. The ANX 1500 provides distinct advantages to electronic design automation (EDA) environments:
- Speeding NFS data delivery to the design process and reducing data latency associated with simulations and verifications. The ANX 1500 seamlessly sits alongside any NAS device and accelerates NFS read performance. The ANX 1500 has the network bandwidth and processing capability to satisfy the extensive simultaneous NFS read requirements characteristic of this environment.
- Accelerating the transport of scripts and libraries to the compute clusters. Because this storage may make use of slower IOPS producing drives, the ANX can greatly reduce latency associated with moving data to the compute farm or clients.
- Improving client performance. The ANX 1500 can handle over 2 million metadata operations per second. When the NAS is no longer burdened with responding to thousands of metadata requests it naturally performs better.
- Saving tremendous amounts of capital, power, space, and cooling expense. By offloading the NAS, far fewer disk drives are required, NAS controllers can scale to support more users and administrators spend less time scaling NAS infrastructure