By Burnie Wu, VP of Business Development, MemVerge
Our company and area of specialization with respect to EDA workloads entails applying memory virtualization and transparent checkpointing of unmodified EDA applications to achieve operational cost savings and/or performance improvements. Hence, our specific predictions for 2023 are based on our unique vantage point into the EDA industry.
- EDA users will continue seeking to migrate workloads to public clouds. At the same time, they will be seeking ways to reduce cloud computing costs through increased use of public spot instances.
- The first generation of CXL memory-enabled x86 servers will begin shipping and be qualified for use with memory-bound EDA workloads. This will lift some of the capacity and bandwidth constraints for EDA applications and improve runtimes.
- Interest in transparent checking pointing/recovery (including rollback) of running EDA applications will grow as implementations mature and more use cases for improving productivity and efficiency are validated by both system administrators and users.
