---
id: "simon-boehm-cuda-matmul-worklog"
title: "How to Optimize a CUDA Matmul Kernel for cuBLAS-like Performance (Simon Boehm)"
creator: null
creator_url: null
creator_x: null
url: "https://siboehm.com/articles/22/CUDA-MMM"
level: "expert"
level_max: "expert"
category: "model-training"
subcategory: "training-infrastructure"
region: "unknown"
license: "unknown"
license_scope: "linked-resource"
access: "free"
models: []
source_date: null
date_added: "2026-10-18"
resource_kind: "learning_resource"
verification_basis: "supplier-inventory-v2.3"
access_evidence: "Supplier-approved free inventory; not independently verified by this site."
domain_top: "Infrastructure, chips & efficiency"
domain_sub: "chips & hardware"
last_verified: "2026-10-18"
---

[Model Training](/model-training/index.md)

# How to Optimize a CUDA Matmul Kernel for cuBLAS-like Performance (Simon Boehm)

Creator not supplied; see original source.

[Read the original →](https://siboehm.com/articles/22/CUDA-MMM)

Simon Boehm's worklog optimising a CUDA matrix-multiplication kernel step by step towards cuBLAS performance.

Level: [Expert](/levels/index.md)
Models: 
Category: [Model Training](/model-training/index.md)
Tags: [chips-hardware](/tags/chips-hardware/index.md), [infrastructure-chips-efficiency](/tags/infrastructure-chips-efficiency/index.md)
Source date: date not stated
Added: 2026-10-18
Last verified: 2026-10-18
Status: ok
Prerequisites: source
Subcategory: [training infrastructure](/model-training/training-infrastructure/index.md)
Region: unknown
Original language: en
Access: free
Resource reuse: unknown: Not supplied; no reuse grant asserted. [Facet definitions](/about/index.md#facets)
Verification basis: Supplier-reported live check; no independent HTTP verification by this site.
Access evidence: Supplier-approved free inventory; not independently verified by this site.
Source domain: Infrastructure, chips & efficiency
Source subdomain: chips & hardware
Supplier region label: Germany

Summary in our own words. Rights remain with the original creator. [Credit and correction policy](/about/index.md#credit-policy).

[Plain-text Markdown copy](/model-training/training-infrastructure/simon-boehm-cuda-matmul-worklog/index.md)