Mnist Neural Network Implemented in Pure X86 Assembly From Scratch

Posted2 months agoActive2 months ago

mghaderi

5 points

3 comments

github.comTechstory

calmpositive

Debate

20/100

Neural NetworksAssembly LanguageMachine Learning

Key topics

Neural Networks

Assembly Language

Machine Learning

A developer has implemented a neural network for MNIST dataset classification in pure x86 assembly language, sparking interest and discussion about the project's technical merits and potential applications.

Snapshot generated from the HN discussion

Discussion Activity

Light discussion

First comment

N/A

Peak period

0-1h

Avg / period

Key moments

01Story posted
Nov 2, 2025 at 12:36 PM EST
2 months ago
Step 01
02First comment
Nov 2, 2025 at 12:36 PM EST
0s after posting
Step 02
03Peak activity
3 comments in 0-1h
Hottest window of the conversation
Step 03
04Latest activity
Nov 2, 2025 at 1:25 PM EST
2 months ago
Step 04

Generating AI Summary...

Analyzing up to 500 comments to identify key contributors and discussion patterns

Discussion (3 comments)

Showing 3 comments

mghaderiAuthor

2 months ago

1 reply

I implemented a neural network from scratch in x86 assembly (no frameworks, no Python) to recognize handwritten digits from MNIST. Feedback on performance optimizations or next steps is welcome Uses AVX-512 SIMD for parallel float32 ops (~7× faster than NumPy). Runs in a lightweight Debian Slim Docker container. The goal was to understand neural networks at the CPU level.

checker659

2 months ago

1 reply

> ~7× faster than NumPy

Is that on the CPU (not sure if NumPy has a GPU backend)

mghaderiAuthor

2 months ago

Yes CPU same resources And same implementation

View full discussion on Hacker News

ID: 45791969Type: storyLast synced: 11/17/2025, 7:49:18 AM

Want the full context?