---
title: "Discussion on \"Why cudaMalloc fails on NVIDIA Jetson Orin Nano Super — and the one flag that fixes it\" | Hashnode"
description: "Discussion on \"Why cudaMalloc fails on NVIDIA Jetson Orin Nano Super — and the one flag that fixes it\". A unified-memory allocator fix for running llama.cpp on Jetson's shared GPU/CPU memory pool."
url: "https://hashnode.com/posts/why-cudamalloc-fails-on-nvidia-jetson-orin-nano-super-and-the-one-flag-that-fixes-it/6a32ada98db3d147576cdde4"
---

[Feed](https://hashnode.com/feed) /

Discussion

# [Why cudaMalloc fails on NVIDIA Jetson Orin Nano Super — and the one flag that fixes it](https://hemkesh.hashnode.dev/why-cudamalloc-fails-on-nvidia-jetson-orin-nano-super-and-the-one-flag-that-fixes-it?utm_source=hashnode&utm_medium=feed)

A unified-memory allocator fix for running llama.cpp on Jetson's shared GPU/CPU memory pool.

Jun 17 · 2 min read · hemkesh.hashnode.dev  
Updated Oct 16  
Read the article

-   [#jetson-orin-nano](https://hashnode.com/tag/jetson-orin-nano)
-   [#nvidia](https://hashnode.com/tag/nvidia)
-   [#cpp-tutorial](https://hashnode.com/tag/cpp-tutorial)
-   [#edge-ai](https://hashnode.com/tag/edge-ai)
-   [#llama-cpp](https://hashnode.com/tag/llama-cpp)
-   [#embedded-systems](https://hashnode.com/tag/embedded-systems)
-   [#deep-learning](https://hashnode.com/tag/deep-learning)
-   [#gpu-computing](https://hashnode.com/tag/gpu-computing)

## Written by

[Hemkesh](https://hashnode.com/@hemkesh)

@hemkesh · Bengaluru, India

  
[View profile](https://hashnode.com/@hemkesh)

Edge AI on Jetson. Computer vision. Real bugs, real fixes.

## Responses

Markdown supported · ⌘↵ to post

No responses yet.