Tutoring LLM into a Better CUDA Optimizer
The result's identifiers
Result code in IS VaVaI
<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216208%3A11320%2F25%3A10501246" target="_blank" >RIV/00216208:11320/25:10501246 - isvavai.cz</a>
Result on the web
<a href="https://link.springer.com/chapter/10.1007/978-3-031-99857-7_18" target="_blank" >https://link.springer.com/chapter/10.1007/978-3-031-99857-7_18</a>
DOI - Digital Object Identifier
<a href="http://dx.doi.org/10.1007/978-3-031-99857-7_18" target="_blank" >10.1007/978-3-031-99857-7_18</a>
Alternative languages
Result language
angličtina
Original language name
Tutoring LLM into a Better CUDA Optimizer
Original language description
Recent leaps in large language models (LLMs) caused a revolution in programming tools (like GitHub Copilot) that can help with code generation, debugging, and even performance optimization. In this paper, we focus on the capabilities of the most recent reasoning models to generate optimized CUDA code for predefined, well-known tasks. Our objective is to determine which types of code optimizations and parallel patterns the LLMs can perform by themselves and whether they can be improved by tutoring (providing more detailed hints and guidelines in the prompt). The generated solutions were evaluated both automatically (for correctness and speedup) and manually (code reviews) to provide a more detailed perspective. We also tried an interactive approach where the LLM can fix its previous mistakes within a session. The results indicate that LLMs are quite skilled coders; however, they require tutoring to reach optimized solutions provided by parallel computing experts.
Czech name
—
Czech description
—
Classification
Type
D - Article in proceedings
CEP classification
—
OECD FORD branch
10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)
Result continuities
Project
—
Continuities
S - Specificky vyzkum na vysokych skolach
Others
Publication year
2025
Confidentiality
S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů
Data specific for result type
Article name in the collection
Euro-Par 2025: Parallel Processing
ISBN
978-3-031-99857-7
ISSN
0302-9743
e-ISSN
1611-3349
Number of pages
14
Pages from-to
250-263
Publisher name
Springer Nature
Place of publication
Cham
Event location
Dresden, Germany
Event date
Aug 25, 2025
Type of event by nationality
WRD - Celosvětová akce
UT code for WoS article
001579606700018