# ParaView 6.1.0 severe UI lag with NVIDIA driver 595 — ?? Qt 6.9 interaction

**URL:** https://discourse.paraview.org/t/paraview-6-1-0-severe-ui-lag-with-nvidia-driver-595-qt-6-9-interaction/17570
**Category:** ParaView Support
**Created:** [May 19, 2026, 10:12am UTC](https://discourse.paraview.org/t/paraview-6-1-0-severe-ui-lag-with-nvidia-driver-595-qt-6-9-interaction/17570 "2026-05-19T10:12:30Z")
**Posts on this page:** 1
**Page:** 1

<div class="post-metadata">

### Author: ![karl.randle](https://discourse.paraview.org/letter_avatar_proxy/v4/letter/k/4da419/32.png) [@karl.randle](https://discourse.paraview.org/u/karl.randle)
#### Post date: [May 19, 2026, 10:12am UTC](https://discourse.paraview.org/t/paraview-6-1-0-severe-ui-lag-with-nvidia-driver-595-qt-6-9-interaction/17570/1 "2026-05-19T10:12:30Z")

</div>

## Summary

ParaView 6.1.0 is extremely slow on a high-end workstation with NVIDIA driver 595.97 (RTX PRO 4000 Blackwell). The slowness affects toggling visibility of pipeline items, switching representations (e.g. outline to surface), and general UI responsiveness. ParaView 5.13.3 on the same machine, same driver, and same data is fast. The issue is therefore assumed to be isolated to the Qt 6.9 layer introduced in ParaView 6.x.

This is maybe related to the previously reported issue: [ParaView not working with NVIDIA driver 595](https://discourse.paraview.org/t/paraview-not-working-with-nvidia-driver-595/17474)

## Diagnostic findings

The Timer Log confirms that ParaView’s internal rendering pipeline is **not** the bottleneck. All render times are normal:

```auto
Still Render, 0.003 seconds
    Render (use_lod: 0), (use_distributed_rendering: 0), (use_ordered_compositing: 0)
    OpenGL Dev Render, 0.001 seconds
    OpenGL Dev Render, 0 seconds
    OpenGL Dev Render, 0 seconds
    OpenGL Dev Render, 0.001 seconds

```

Despite these fast internal render times, the perceived UI response when toggling pipeline visibility or switching to surface representation is on the order of seconds — indicating the delay occurs between ParaView completing its OpenGL render and the frame being presented on screen.

## Environment

**ParaView 6.1.0 (slow - like ridiculous to change pipelines):**

```auto
Version: 6.1.0
VTK Version: 9.6.1
Qt Version: 6.9.0
SMP Backend: TBB
SMP Max Number of Threads: 32
OpenGL Vendor: NVIDIA Corporation
OpenGL Version: 3.2.0 NVIDIA 595.97
OpenGL Renderer: NVIDIA RTX PRO 4000 Blackwell/PCIe/SSE2
MPI Enabled: Off
ParaView Build ID: superbuild af9bb3207d40b64f97c64a9c319d483ff3b17d03 (!1332)

```

**ParaView 5.13.3 (lightning fast on the same system):**

- Qt 5.15, same hardware, same NVIDIA driver 595.97

- All operations responsive, no perceptible lag

**Hardware:**

- CPU: AMD Ryzen 9 9950X (16C/32T)

- RAM: 96 GB DDR5-5600 ECC

- GPU: NVIDIA RTX PRO 4000 Blackwell (it does not even get above 2%)

- OS: Windows 11 Pro

- Single monitor, no remote desktop

## Reproduction

1. Install ParaView 6.1.0 on a system with NVIDIA driver 595.xx

2. Open any dataset with multiple pipeline items

3. Toggle visibility on/off in the pipeline browser, or switch representation from outline to surface

4. Observe multi-second UI lag despite fast Timer Log render times

5. Repeat the same workflow in ParaView 5.13.3 — lag is absent

## Conclusion

The performance regression is in the Qt 6.9 composition/presentation layer when running on NVIDIA driver branch R595. The VTK rendering pipeline and GPU are performing normally. Rolling back to ParaView 5.13.3 (Qt 5.15) resolves the issue without any driver changes.

Question is - how to solve this on the newer Paraview releases….. #not an Ai but did get help from Claude.