Release Notes

VPI-4.0.7

This VPI version supports:

• Jetson AGX Thor DevKit
• Jetson AGX Thor T5000 and T4000
• IGX Thor DevKit + Blackwell RTX6000 dGPU
• IGX Thor T7000 and T5000
• IGX Thor Safety DevKit

Linux x86_64 with NVIDIA dGPUs

• CUDA 12: Maxwell or newer (sm_52 or newer)
• CUDA 13: Volta or newer (sm_70 or newer)
• tested with Ubuntu 24.04

For previous release notes refer to Previous Releases

Key Highlights

dGPU on IGX Platform Support

• Blackwell RTX PRO 6000 (sm_120) — Added CUDA sm_120 architecture support for the NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition on IGX. This GPU was not previously included in the Tegra CUDA gencode flags.
  • Changed: CMakeLists.txt — added sm_120 to the -gencode list alongside existing sm_87, sm_103, sm_110, sm_121.

dGPU on IGX: Known Limitations

• OFA not available on dGPU — The Optical Flow Accelerator (OFA) is not supported by VPI on dGPU. Algorithms that rely on OFA are not supported on dGPU on IGX, including Dense Optical Flow (vpiSubmitOpticalFlowDense) and Stereo Disparity when using the OFA backend.
• No shallow copies on dGPU — On dGPU the GPU has its own discrete memory, so zero-copy (shallow copy) between host and GPU is not possible. All data transfers require explicit deep copies. VPI objects should be created with VPI_BACKEND_CUDA | VPI_BACKEND_CPU flags.

Bug Fix: LK Optical Flow CPU

• Fixed feature tracking degradation over time in CPU LK Optical Flow (vpiSubmitOpticalFlowPyrLK with VPI_BACKEND_CPU). A memory mapping issue caused the tracking status array to accumulate stale values, progressively marking valid features as failed. Feature retention is now stable (~190-200 features vs. dropping to 6 previously).

---

Compatibility

• API: Fully compatible with VPI 4.0.0
• ABI: Maintained across all 4.0.x releases
• Platform: All VPI 4.0.x supported platforms
• Requirements: BSP 38.2.1-PRC2+, CUDA 13.0

---

Known Issues Resolved (since v4.0.5)

• LK Optical Flow CPU: Feature tracking degradation over time

---

Known Issues

• StereoDisparity: The following block linear formats are not working: VPI_IMAGE_FORMAT_NV12_ER_BL, VPI_IMAGE_FORMAT_NV12_ER_BL16, VPI_IMAGE_FORMAT_NV12_BL, VPI_IMAGE_FORMAT_NV12_BL16. Workaround is to convert those image formats.
• Dense Optical Flow: There is no discernible difference in quality from low, medium to high.
• dGPU on IGX: OFA not available. Dense Optical Flow and OFA-based Stereo Disparity are not supported.
• dGPU on IGX: No shallow copies. Use VPI_BACKEND_CUDA | VPI_BACKEND_CPU flags.
• dGPU on IGX: Corner detection samples (Harris Corners, ORB) produce corrupted artifacts on CUDA backend. Workaround: create VPI objects with VPI_BACKEND_CUDA | VPI_BACKEND_CPU flags.
• Background Subtractor: Initial frames may produce unstable background model.
• PyTorch: 16-vpi_pytorch sample fails due to the absence of a compatible PyTorch module for CUDA Toolkit 13.0.
• ORB: pva backend does only produce FAST keypoints.
• Jetson Thor - VIC, Interpolation Limitation: VPI_INTERP_NEAREST not available.

Notices

Disclaimer

ALL NVIDIA DESIGN SPECIFICATIONS, REFERENCE BOARDS, FILES, DRAWINGS, DIAGNOSTICS, LISTS, AND OTHER DOCUMENTS (TOGETHER AND SEPARATELY, "MATERIALS") ARE BEING PROVIDED "AS IS." NVIDIA MAKES NO WARRANTIES, EXPRESSED, IMPLIED, STATUTORY, OR OTHERWISE WITH RESPECT TO THE MATERIALS, AND EXPRESSLY DISCLAIMS ALL IMPLIED WARRANTIES OF NONINFRINGEMENT, MERCHANTABILITY, AND FITNESS FOR A PARTICULAR PURPOSE.

Information furnished is believed to be accurate and reliable. However, NVIDIA Corporation assumes no responsibility for the consequences of use of such information or for any infringement of patents or other rights of third parties that may result from its use. No license is granted by implication of otherwise under any patent rights of NVIDIA Corporation. Specifications mentioned in this publication are subject to change without notice. This publication supersedes and replaces all other information previously supplied. NVIDIA Corporation products are not authorized as critical components in life support devices or systems without express written approval of NVIDIA Corporation.

Trademarks

NVIDIA and the NVIDIA logo are trademarks or registered trademarks of NVIDIA Corporation in the U.S. and other countries. Other company and product names may be trademarks of the respective companies with which they are associated.

Copyright

© 2019-2025 NVIDIA Corporation. All rights reserved.