FAQ - Frequently Asked Questions

Here you'll find answers to common questions about using the Wattlytics tool. If your question isn’t listed, feel free to use the feedback form.

General Questions

What is Wattlytics?

It’s a cost-efficiency metric calculated as: GPU performance ÷ Total Cost of Ownership over the system lifetime. It helps identify which GPUs deliver the most performance per euro spent.

Who should use this tool?

It’s useful for researchers, HPC administrators, grant planners, and procurement teams evaluating GPU nodes based on real budget constraints and workload needs.

Technical Documentation

Do I need to fill out every slider?

No. Default values are provided based on typical HPC cluster assumptions. You can modify only the parameters relevant to your scenario.

What does “Same number of GPUs for all types” mean?

When enabled, each GPU type is compared using the same number of GPUs per node. When disabled, the configuration scales GPUs to fit the budget.

How are benchmark IDs used?

Each ID corresponds to a measured or modeled performance and power for GROMACS or AMBER workloads. Select the one closest to your target use case.

Can I export the results?

Yes. You can export both the performance/TCO results and sensitivity analysis as CSV files and charts for further analysis or reporting.

Tip Features & Blog Generator

How does the Tip Assistant work?

As you adjust inputs, the assistant provides smart suggestions based on current values and cost-impact heuristics.

What is the auto blog generator?

After running a calculation, you can generate a Markdown-style summary of the analysis — ideal for documentation, proposals, or sharing insights.

Privacy & Data

Is any data saved or tracked?

No user input is stored or sent anywhere. All calculations are performed locally in your browser.

Is the tool open source?

Only partially — but it may be released under a permissive license in the future. Stay tuned!