Advanced Modeling & Simulation
Advanced Modeling & Simulation builds on HPC Foundations with the tools to automate, organize, and reuse your simulation work.
Workflows chain jobs, workstations, and decisions into a template you configure once and run again. Automations run your own scripts before, during, or after a job, such as one that writes results into searchable custom fields.
Rescale Edison answers questions about your jobs, files, and spend in plain language. Three agents help you choose hardware, understand failed runs, and track costs. Data Connectors bring files from AWS S3 and Azure Blob Storage into your jobs without storing a copy in Rescale.
What is a workflow?
A workflow template is a graph of connected steps that you configure once and run as often as you need with new inputs. Each step is a node: a job, a workstation, a storage device, or a decision that routes the run based on results. Builders lay out the template on a visual canvas, and anyone it's shared with can run it by supplying the inputs.
Every run keeps a record of what ran and with which inputs. If a step fails, a retry re-runs only the failed, changed, and downstream steps instead of starting over.
Benefits
- Run multi-step processes the same way every time. Build a workflow once on a visual canvas, and anyone it's shared with runs it by supplying inputs.
- Capture results without manual entry. Automations write metrics such as residuals and runtime into custom fields on every run.
- Find and compare past work. Search by tags and custom fields, and report on them across jobs and in billing exports.
- Choose hardware on evidence. Benchmark your own workload across coretypes and compare cost against time to solve.
Key capabilities
Advanced Modeling & Simulation includes everything in HPC Foundations and adds the capabilities below.
Workflows
Chain jobs, workstations, storage devices, and decisions into a template on a visual canvas, with no scripting. Outputs pass between steps through shared storage, and a decision step can route the run, for example to different hardware. Start from a blank canvas or a workflow template from Rescale.
Use it to: standardize a multi-step process so everyone on the team runs it the same way.
Automations
Attach your own Python or Bash script to a job to run before, during, or after the solver. The most common use is extracting metrics such as convergence statistics, residuals, and runtime into custom fields.
Use it to: capture results on every run, post completion messages to Slack or Microsoft Teams, or convert and copy output files.
Tags & Custom Fields
Tags are free-form labels for finding jobs, workstations, and files. Custom fields are typed values with a defined scope, and a required field blocks submission until it's filled in. Custom fields also appear as columns in billing reports.
Use it to: search, compare, and report on your work by the values that matter to your program.
Data Connectors
Browse AWS S3 and Azure Blob Storage from inside Rescale and attach files as job inputs. Files transfer straight to the compute cluster at launch, and Rescale keeps no copy.
Use it to: run on data that's too large to move or has to stay where it is.
Hardware Performance
Compare coretypes side by side on published specs with Coretype Explorer. Then benchmark your own workload across coretypes and core counts with Performance Profiles, plotted as cost against time to solve.
Use it to: choose hardware on measured cost and speed, not specs alone.
Rescale Edison and AM&S agents
Ask Edison about your jobs, files, results, and spend in plain language, on the web platform or in Rescale Desktop. Three agents are included: Hardware Recommender, Job Failure Summarizer, and Finance UI Agent. Each user gets a daily Edison allowance that resets every day.
Use it to: find out why a run failed, which hardware to pick, or where your budget went.
→ Rescale Edison · AM&S agents
How Advanced Modeling & Simulation compares
Each tier includes everything in the one before it. AI Physics is added on top of Advanced Modeling & Simulation or Agentic Engineering, with its own AI Credits pool.
Common use cases
Pre- and post-processing pipelines
A workstation or job step generates the mesh or input files, solver jobs run in parallel across parameter variations, and a final job aggregates the results. Files move between steps automatically.
Route a run to the right hardware
A pre-processing job reports the mesh size, and a decision step sends the solver job to a smaller or larger hardware configuration. Benchmark the candidate configurations with Performance Profiles to set that threshold on measured cost and time.
Capture results on every run
Attach an automation that reads the solver output after each run and writes residuals, convergence statistics, and runtime into custom fields. You can then search and compare runs by those values and see them as columns in billing reports.
Run on data that stays in your cloud storage
Attach input files from AWS S3 or Azure Blob Storage during job setup. They go straight to the compute cluster at launch, so large datasets don't have to be uploaded to Rescale first.
Prerequisites
- Your workspace is on the Advanced Modeling & Simulation tier, or on Agentic Engineering. Your Rescale account team can confirm what your workspace includes.
- To run a workflow, you own the template or it's shared with you.
- To attach files from cloud storage, a workspace admin has set up an AWS S3 or Azure Blob Storage connector.
- To use Edison and agents, you work on a platform where Edison is available. It isn't available on the ITAR or FedRAMP platforms.
Get started
Each step links to its full guide.
- Build a workflow template: lay out a multi-step pipeline on the canvas, or start from a workflow template from Rescale.
- Run a workflow: submit a template with your inputs and follow each step.
- Create an automation: attach a script that captures results into custom fields on every run.
- Ask Rescale Edison: ask about your jobs, a failed run, or your spend in plain language.
FAQs
Does Advanced Modeling & Simulation use AI Credits? No. Each user gets a daily Edison allowance that resets every day, and unused allowance doesn't carry over. Workspaces on this tier don't draw from an AI Credits pool.
What happens when I reach my daily Edison limit? Edison stops answering until your allowance resets the next day, and the usage bar in Edison shows the limit is reached. Jobs, workflows, and everything else keep working. If your team needs more, Agentic Engineering replaces the daily allowance with a shared AI Credits pool.
Do I need to write code? Not for Workflows: you build templates on a visual canvas. Automations are Python or Bash scripts that you attach to a job.
Can agents make changes on their own? No. Anything that changes something, such as submitting a job, asks for your approval first.
How do I get more agents? Agentic Engineering includes the three Advanced Modeling & Simulation agents and adds the full library, including agents that validate input files, fix and resubmit failed jobs, and help build workflows.
Does Rescale use my data to train AI models? No. Rescale doesn't use your prompts, Edison's answers, or your engineering data to train or fine-tune AI models, and doesn't let third parties do so.
Support
Contact your Rescale representative or visit the Rescale Support Center.
Related
- Agentic Engineering agents — the full agent library that Agentic Engineering adds
- AI Physics — surrogate model training you can add to this tier
- AI credits — how AI usage is metered on Agentic Engineering and AI Physics
- HPC Foundations — the jobs, workstations, and storage this tier builds on