Cadence Hybrid Cloud Simplifies Data Lift and Shift

17 Apr 2024 • 5 minute read

The semiconductor industry thrives on innovation, and at the heart of this progress lies Electronic Design Automation (EDA). EDA tools allow engineers to design and evaluate chips, before manufacturing, a data-intensive process. It would not be wrong to say that data is the lifeblood that powers every operation, analysis, and service of EDA design. Further, the EDA design has evolved from siloed/local design to designs across global teams, resulting in geographically distributed compute clusters. Each regional cluster often involves an adjacent data cluster. With the increasing data handling/storage locations, design teams often encounter a formidable barrier – consistent data access and management.

Hybrid Cloud EDA and Data Lift and Shift

Cadence has implemented an industry-leading True Hybrid Cloud solution to mitigate such barriers as more and more customers trying to migrate to Cadence Cloud. The True Hybrid Cloud is an innovative approach for managing data for design teams. Unlike the traditional "lift-and-shift" method of moving the entire design data set to the cloud, the True Hybrid Cloud allows teams to transfer only the necessary portion of their files, typically about 10%, to the Cadence-managed cloud for computation.

This blog discusses how Cadence revolutionizes the landscape for EDA designers by dramatically slashing the time-to-market for new product development, leveraging unused computec in remote data centers and elastic compute from the Cloud. This is only possible with consistent data access and management across these regions.

Cloud Data Conundrum

Transitioning to cloud computing brings a wealth of advantages but also introduces unique challenges, particularly in data management. When we migrate EDA workloads to the cloud, we encounter significant obstacles, including:

Migrating dependent data from on-prem to the cloud or between data centers
Maintaining regular synchronization from on-prem to the cloud or between DC
Managing voluminous data that is not in use 
Retrieving required results or logs from cloud to on-premises infrastructure
Due to the large data volume, time is also a big concern

Such issues can significantly hamper efficiency and escalate costs in terms of time and resources.

From Concept to Reality

The rest of this blog describes an implementation of the Cadence Hybrid Cloud solution for users across the globe.

Users submit jobs in their local HPC cluster; if it has enough cores to process the job, it runs locally; otherwise, the job is forwarded to a remote cluster. The HPC MultiCluster feature connects the local HPC cluster to remote clusters. Remote clusters can be in other data centers with static compute configuration or in the Cloud. Compute on the remote cluster on Cloud is scaled up and down as required using Cadence-built automation. Required results data from the Cloud are fetched back. For the user, their job is running locally. They continue to perform debugging and failure analysis at the local cluster seamlessly.

Cadence also leverages NetApp FlexCache technology to enable customers to access data from anywhere, anytime, without explicitly synchronizing incremental changes. This capability, offered in Cadence OnCloud Managed Service solutions, reduces the initial setup time from days to hours and altogether removes the need for explicit synchronization of incremental data changes, resulting in faster time to market for new product development and introduction and gaining a fast-moving advantage in the industry. This capability is available for customer-managed hybrid cloud environments as well.

Implementing FlexCache architecture follows a systematic approach, starting with the essential network and source configuration step. This phase ensures seamless communication between the storage systems involved in the FlexCache setup. Network configuration mandates proper connectivity, enabling crucial ports if firewalls are present and setting up IP addresses, VLANs, and Intercluster LIF (ICL). Furthermore, cluster peering between on-premises clusters and remote NetApp Clusters is configured, setting a robust foundation for the subsequent creation of FlexCache volumes.

Network Configuration

Ensure proper network connectivity between the NetApp storage systems involved in the FlexCache setup
If the firewall involves between the data centers or public cloud, ensure necessary ports (10000, 11104, 11105, 10566, 10569, 443 (TCP) are enabled
Configure network settings such as IP addresses, VLANs, and Intercluster LIF (ICL)
Configure cluster peering between your on-premises cluster and the Remote NetApp Cluster

Source Configuration

FlexCache configurations enable linking a single NetApp ONTAP Cluster/SVM source volume to as many as 100 destination FlexCache volumes across NetApp ONTAP Clusters/SVMs. These volumes can operate on both on-premises ONTAP systems and cloud-based instances like CVO or FSxN. Establishing peer relationships between the source and destination clusters and among the source and destination storage virtual machines (SVMs) is essential to facilitate seamless data flow.

Cache Configuration

With the setup, creating FlexCache volumes at the destination becomes a streamlined process. These volumes serve as the cache for the data being accessed and help to cater to the data demands and improve the access speeds dramatically. These volumes can be created by

Specifying the source volumes from which the FlexCache volumes will cache data
Set up appropriate export policies and rules to allow FlexCache access to the data

Mounting these volumes on client systems further closes the loop of ensuring that the cached data serves its intended purpose effectively and efficiently. Cadence customers can use NetApp FlexCache in various use cases, including.

Distributed environment: FlexCache can cache data from a primary storage system onto one or more remote caching systems, enabling users in remote locations to access data more quickly
Cloud environment: FlexCache can cache data in a cloud environment, improving data access times and reducing network latency

Conclusion

With the groundbreaking approach to overcoming the traditional challenges of cloud data management, Cadence enables customers to implement a true hybrid cloud solution. It leverages FlexCache to create a cache of frequently accessed data on a remote NetApp storage system, which the client system can access without incurring the overhead of accessing the primary storage system. This can lead to significant performance improvements for workloads that require frequent access to the same data. Further, read and write volumes in the cache are separated for better performance.

Implementing a hybrid cloud solution requires careful planning and design to ensure that the solution meets the specific business requirements and integrates with the existing infra. Essential factors to consider include job routing logic, FlexCache nodes, the size of the FlexCache volume, the connectivity between source and destination locations, and egress costs, if any.