Keywords
Abstract
This article explores the application of segmented-paged memory management in artificial intelligence and machine learning, particularly in convolutional neural networks (CNN). It analyzes the memory demands of CNN models, GPU architecture, CUDA Unified Memory technology, and the efficiency of segmentation and paging in AI systems. Practical examples demonstrate how memory management approaches can serve as effective optimization tools in real-world projects.
References
1. GeeksforGeeks. (2022). Paged Segmentation and Segmented Paging. https://www.geeksforgeeks.org/paged-segmentation-and-segmented-paging/
2. NVIDIA Developer Blog. (2021). Unified Memory for CUDA Beginners. https://developer.nvidia.com/blog/unified-memory-cuda-beginners
3. NVIDIA Developer Blog. (2022). Maximizing Unified Memory Performance in CUDA. https://developer.nvidia.com/blog/maximizing-unified-memory-performance-cuda
4. arXiv.org. (2024). Memory-Efficient Training of CNNs with Input Segmentation. https://arxiv.org/abs/2408.03663
5. Madumarov U. A., Abdurahmonov A. O. (2019). Operatsion tizimlar. – Toshkent: “Fan va texnologiya” nashriyoti.
6. Abdullayev X. M., Karimov A. A. (2020). Sun’iy intellekt va mashinaviy o‘qitish. – Toshkent: “Innovatsiya” nashriyoti.
7. Qurbonov B. Q., Eshqobilov A. T. (2021). Parallel hisoblash tizimlari. – Samarqand: SamDU nashriyoti.
8. Tutorialspoint. (2023). Memory Management in Operating Systems. https://www.tutorialspoint.com/operating_system/os_memory_management.htm
9. NVIDIA CUDA Toolkit Documentation. (2024). Unified Memory Programming Guide. https://docs.nvidia.com/cuda/cuda-c-programming-guide/index.html#um-overview
10. O‘zbekiston Respublikasi Oliy va o‘rta maxsus ta’lim vazirligi. (2018). Kompyuter grafikasi va dasturlash: o‘quv qo‘llanma. – Toshkent: “Yangi asr avlodi”.
11. Umarov, B., G'ulomjonova, S. (2024). BULUT TEXNOLOGIYASI VA ULARDAN FOYDALANISH.