In the contemporary landscape of computational lithography, the demand for efficient utilization of compute resources has surged. The escalating complexity of models and simulations for advanced nodes necessitates a strategic approach to maximizing throughput while maintaining high resource utilization.
For over a decade, Calibre Cluster Manager (CalCM) has been pivotal in enabling customers to optimize their compute clusters' usage throughout post-tapeout flow operations. By facilitating automatic resource allocation, it significantly enhances turnaround time across various workload tasks, from retargeting to OPCVerify.
Now, a new era emerges with the introduction of CalCM+, which integrates artificial intelligence to revolutionize post-tapeout flow operations for both on-premises and cloud environments. Through the fusion of analytics and observability, it achieves cost optimization while ensuring optimal resource allocation.
The incorporation of machine learning-based runtime prediction and advanced resource control empowers users to effectively rightsize their on-premises and cloud resources. Moreover, advanced monitoring and outlier detection systems mitigate critical issues within complex clusters throughout the tapeout process, enhancing overall operational reliability and efficiency.