Open Issues Need Help
View All on GitHubAI Summary: Optimize the traccc project's doublet finding algorithm by pre-computing linearised circles within the `find_doublets` kernel, aiming to improve performance by merging existing kernels and deduplicating computations between mid-bottom and mid-top doublets. This requires modifying the CUDA code to integrate the pre-computation step efficiently within the existing kernel.
Complexity:
4/5
good first issue performance