NVIDIA · Python DLA Runtime
cuda.bindings.cudla (cuDLA Python API)
NVIDIA has added cuda.bindings.cudla, a Cython-generated Python layer over the cuDLA C++ API, introduced in cuda-bindings 13.3.0 as part of the CUDA Python project. It exposes thirteen functions, nine data types and eight enums for hybrid-mode workflows, where CUDA streams coordinate DLA inference with GPU preprocessing and postprocessing on Jetson AGX Orin. The post walks through a nine-step image classification lifecycle, and the API is stated to be identical on NVIDIA RTX Spark Windows laptops. Standalone mode is not supported in this release.
Open the source ↗Similar Developer Tools releases
- llama.cpp v1.0.0
ggml-org and Hugging Face · Local inference library · Week 47, 2026
- MACaQ
Huawei CSL (huawei-csl) · Model quantisation toolkit · Week 47, 2026
- Needle 3
Cactus Compute · Tiny Device Model · Week 39, 2026
- edge0
Edge0-AI · Streaming MoE Inference · Week 37, 2026
- Magnitude
magnitudedev · Local Model Server · Week 36, 2026
- SwarmLLM
SwarmLLM · Peer-to-Peer Inference · Week 36, 2026
More from NVIDIA
- cuBLAS Ozaki-II and 2M FP64 Algorithms
NVIDIA · FP64 GEMM Emulation · Week 47, 2026
- NVIDIA SOMA Retargeter v0.2
NVIDIA · Motion Retargeting Tool · Week 47, 2026
- Agent Substrate
NVIDIA · Agent Sandbox Runtime · Week 47, 2026
- NVIDIA OpenShell, Snapshot and Kubernetes Engine
NVIDIA · Agent Runtime Infra · Week 47, 2026
- NVIDIA Alpamayo 2 Edge and AlpaJudge
NVIDIA · Driving VLA model · Week 47, 2026
- GR00T 2
NVIDIA · Open World Action Model · Week 47, 2026
Published by Quanta AI Labs, an AI agency in India. Browse every week in AI News Weekly.