TotalCAE clients are increasingly utilizing TotalCAE Hybrid HPC to increase flexibility for engineering simulation for the following five major scenarios:
- To meet simulation demand beyond on-premise capacity to win new business.
- To offload large CFD simulation and Design of Experiments (DOE).
- Eliminate queue time for hot jobs that need fast turnaround to impact product design.
- To place HPC resources closer to remote engineering teams.
- As a disaster recovery site for on-premise HPC clusters.
This post will dig into these usage scenarios.
Winning new business
Engineering teams often need to respond to request for proposals with fast turnaround times, that require additional HPC capacity that had not been planned for and are beyond the capacity available on-premise. Often adding additional capacity is difficult to procure quickly enough to respond to the proposal, and capex funding difficult to obtain because the proposals are not guaranteed new business. Utilizing a hybrid HPC strategy to rent additional public cloud hardware capacity to be able to respond to the proposals has enabled TotalCAE clients to win millions of dollars of new business that would have been not possible if constrained by on-premise HPC capacity.
Offload large scale simulation
Many TotalCAE clients run large scale simulations that need thousands of cores, but only in the up-front part of the design cycle. This type of large HPC capacity requirements may only last for a few months, and so customers do not want to purchase on-premise HPC capacity that can’t be fully utilized outside this up-front design cycle. A common use case is offloading Computation Fluid Dynamic (CFD) design by utilizing Hybrid HPC to send CFD onto the public cloud to run thousands of CPU to avoid increasing on-premise wait times during these busy peaks in the design life cycle.
Eliminate queue time for hot jobs
On-premise systems have limited hardware capacity, and engineering teams often have to manage workloads when an on-premise system does have enough capacity to ensure time sensitive jobs run before less important jobs. With hybrid HPC, hot jobs can immediately run without hardware queue time. Engineers utilizing hybrid HPC gain an immediate productivity boost to meet their simulation deadlines that they would not have been able to achieve without access to the TotalCAE hybrid HPC.
Looking at HPC system utilization provided by TotalCAE Analyzer, clients can now use hybrid HPC to meet peak demands when their on-premise system is highly utilized, to avoid queue times that delay engineering getting time sensitive results back to product teams.
Use Hybrid HPC to Meet Peak Demands
Move resources closer to remote engineering teams.
Product engineering teams typically have a core engineering group located near where the on-premise HPC system is located. Some smaller specialized engineering teams in remote locations have difficulty accessing the on-premise system but still could benefit from better HPC access. With a hybrid HPC solution, compute resources can be me made available to these remote teams on-demand as needed, with closer data centers to them to reduce time waiting to download from remote HPC systems.
Disaster recovery for on-premise HPC systems.
It is often cost prohibitive to have a duplicate HPC cluster on-premise for disaster recovery purposes. A Hybrid HPC solution enables a cost-effective cluster “in the cloud” as a backup to the primary on-premise HPC system to keep engineers working productively in the event of a disaster.
TotalCAE Hybrid HPC
TotalCAE Hybrid HPC enables customers to enjoy the benefits of an on-premise HPC appliance from TotalCAE with flexible compute power provided by the TotalCAE cloud infrastructure.
By combining the best qualities of on-premise HPC computing and flexible pay-as-you-go rentable infrastructure, our clients reduce their time to market and meet ever shrinking product design life cycles with a flexible hybrid HPC system managed by TotalCAE.