---
title: "Discussion on \"CUDA for Wave Simulation - Part 3: Efficient CUDA\" | Hashnode"
description: "Discussion on \"CUDA for Wave Simulation - Part 3: Efficient CUDA\". Non-trivial CUDA\nThe current snippet of code calling CUDA is:\ncuda_step<<< 1, 1 >>>\n\nThis uses only one CUDA thread, and is probably extremely inefficient. To verify this, let’s add some way to time t"
url: "https://hashnode.com/posts/efficient-cuda/68b45a9bfd8b445cf425b1a6"
---

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

Discussion

# [CUDA for Wave Simulation - Part 3: Efficient CUDA](https://pbhnblog.ballif.eu/efficient-cuda?utm_source=hashnode&utm_medium=feed)

Non-trivial CUDA The current snippet of code calling CUDA is: cuda\_step<<< 1, 1 >>> This uses only one CUDA thread, and is probably extremely inefficient. To verify this, let’s add some way to time t

Aug 31, 2025 · 9 min read · pbhnblog.ballif.eu  
Updated Aug 11  
Read the article

-   [#cuda](https://hashnode.com/tag/cuda)
-   [#cpp](https://hashnode.com/tag/cpp)
-   [#benchmarking](https://hashnode.com/tag/benchmarking)

## Written by

[Pierre Ballif](https://hashnode.com/@warggr)

@warggr

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

## Responses

Markdown supported · ⌘↵ to post

No responses yet.
