JavaScript Gantt for Large Datasets
“Supports thousands of tasks” is not a useful performance guarantee by itself. A large Gantt can be cheap or expensive depending on hierarchy depth, dependency density, calendar rules, resource leveling, custom cell complexity, and how often the application replaces data.
The two performance problems in a Gantt
First, the browser must render and update an editable task table and a time-scaled chart. RevoGrid addresses the task-table side with row and column viewport virtualization, so the full dataset is not mounted as DOM cells.
Second, the scheduling engine must calculate dates, dependencies, constraints, progress, warnings, and resources. The checked-in RevoGrid Gantt benchmark suite includes two explicit workloads:
| Benchmark case | Included work |
|---|---|
| 1,000 tasks | Summaries, dependencies, resources, warnings, holidays, dependency clamping, and progress rescheduling |
| 3,000 tasks | Automatic resource leveling plus calendar and dependency scheduling |
These scenarios are repeatable engineering evidence, not a promise that every 3,000-task project reaches a particular FPS. The current implementation notes also distinguish existing RevoGrid virtualization from planned Gantt-specific server/window models.
How to benchmark your project
Use production-shaped data, not evenly spaced synthetic bars. Measure:
- initial data load and first usable interaction;
- vertical and horizontal scroll stability;
- hierarchy expand/collapse cost;
- task drag, resize, and dependency propagation latency;
- critical-path and baseline overlay updates;
- resource diagnostics or leveling cost;
- custom renderer and editor churn;
- memory after repeated project changes.
Also test slower client hardware and the browsers your customers actually use. For very large portfolios, consider project partitioning, lazy-loading detail, and server-owned summaries even when the UI is virtualized.
Open the live Gantt demo for interaction quality, then request a Pro Advanced trial to profile your own project. See critical path, resource scheduling, and the full RevoGrid Gantt performance evidence.
Large-data Gantt FAQ
Can RevoGrid Gantt handle thousands of tasks?
The architecture and benchmark cases are designed for large workloads, but your real capacity must be tested with your scheduling rules and product customizations.
Is virtualization enough?
No. It reduces rendering work; scheduling graphs, resource calculations, network transfer, and application state updates still need independent measurement.