---
title: "NVIDIA Video Codec SDK 13.1: Zero-Copy Transcode, AV1 B-Frames, and Frame-Accurate Seek"
description: "Modern video pipelines require higher efficiency and support for complex workloads across streaming, collaboration, and artificial intelligence media tools. NVIDIA released Video Codec SDK 13.1 to expand GPU-accelerated encoding, decoding, and transcoding capabilities. The release introduces Hierarchical Reference Mode for AV1 with up to 31 B-frames, combines UHQ tuning with iterative encoding, and extracts per-macroblock decode statistics for H.264 and HEVC without CPU overhead. It also implements frame-accurate, GOP-aware seeking through the NvVideoDecoder class and adds stereoscopic MV-HEVC decode metadata. To simplify setup across different environments, the update delivers an official multi-stage Docker container pinned to Ubuntu 22.04 LTS, CUDA 12.3.2, and Vulkan SDK."
---

# NVIDIA Video Codec SDK 13.1: Zero-Copy Transcode, AV1 B-Frames, and Frame-Accurate Seek

[NVIDIA](https://yomu.fyi/company/nvidia-developer-blog) · Elizabeth Goodman · Jul 31, 2026

**Type:** Announcement

## Summary

Modern video pipelines require higher efficiency and support for complex workloads across streaming, collaboration, and artificial intelligence media tools. NVIDIA released Video Codec SDK 13.1 to expand GPU-accelerated encoding, decoding, and transcoding capabilities. The release introduces Hierarchical Reference Mode for AV1 with up to 31 B-frames, combines UHQ tuning with iterative encoding, and extracts per-macroblock decode statistics for H.264 and HEVC without CPU overhead. It also implements frame-accurate, GOP-aware seeking through the NvVideoDecoder class and adds stereoscopic MV-HEVC decode metadata. To simplify setup across different environments, the update delivers an official multi-stage Docker container pinned to Ubuntu 22.04 LTS, CUDA 12.3.2, and Vulkan SDK.

## Context

Video pipelines need faster and more efficient processing to handle demanding use cases like immersive streaming, generative AI tools, and content delivery, while traditional SDK setup requires manually compiling samples and managing dependencies across host operating systems and drivers.

## Approach / What changed

NVIDIA updated Video Codec SDK to version 13.1, adding hierarchical AV1 B-frame reference structures, iterative encoding with UHQ tuning, GPU-resident macroblock decode statistics, a GOP-aware frame seeking API, and an official two-stage Docker development image.

## Takeaways

- Hierarchical Reference Mode for AV1 supports up to 31 B-frames arranged in a tree structure to improve compression quality without performance penalties, though video memory usage increases.
- The NVDECODE API exposes hardware-extracted per-macroblock decode statistics—including QP, coding unit type, and motion vectors—for H.264 and HEVC into a GPU buffer with zero CPU overhead.
- The GOP-aware seek implementation in NvVideoDecoder seeks to the nearest preceding IDR frame, skips non-reference frames, and bypasses post-processing mapping until the target frame is reached.

**Tags:** [Docker](https://yomu.fyi/topic/docker), [Performance](https://yomu.fyi/topic/performance), [Streaming](https://yomu.fyi/topic/streaming)

- Source: [NVIDIA](https://developer.nvidia.com/blog/nvidia-video-codec-sdk-13-1-zero-copy-transcode-av1-b-frames-and-frame-accurate-seek)
- Source URL: https://developer.nvidia.com/blog/nvidia-video-codec-sdk-13-1-zero-copy-transcode-av1-b-frames-and-frame-accurate-seek
- Ingested by Yomu: 2026-08-28T00:00:58.574Z

[Read original post](https://developer.nvidia.com/blog/nvidia-video-codec-sdk-13-1-zero-copy-transcode-av1-b-frames-and-frame-accurate-seek)