Cloud native EDA tools & pre-optimized hardware platforms
Verification makes up between 60-80 percent of the compute workload for companies today. Today, companies run verification workloads with high utilization rates, but faster time-to-results and higher quality-of-results are driving them to use public clouds for both burst and sustained capacity.
These requirements come with their own set of unique challenges. We can categorize the challenges into each of the layers of the traditional infrastructure architecture for verification, comprised of:
You can learn more about 草榴社区’ cloud initiatives here.
This video focuses on the challenges at the compute layer and some of the efforts 草榴社区 is doing to help our customers use public cloud resources more cost effectively by partnering closely with all the major public cloud vendors and data storage providers.
This video focuses on the challenges of the data storage layer. While traditional EDA infrastructure primarily uses NFS for sharing of data across all the compute nodes, it performs poorly when the data has to be shared. This can present a challenge when using public clouds to handle bursting needs.
In conclusion, the shift towards using public clouds for handling verification workloads comes with unique challenges, particularly at the compute and data storage layers. 草榴社区 is actively addressing these challenges, partnering with major public cloud vendors and data storage providers, to facilitate cost-effective utilization of these resources. Despite the hurdles, such as poor performance of NFS in the traditional EDA infrastructure, the drive towards achieving faster results and higher quality continues to influence the move towards the use of public clouds for both burst and sustained capacity.
Verify the entire SoC with industry-leading VCS? simulation, Verdi? debug, VC SpyGlass? RTL static signoff, VC Formal? Apps, and silicon-proven Verification.