Nvidia CUDA Explained - C/C++ Syntax Analysis and Concepts

Поделиться
HTML-код
  • Опубликовано: 15 июл 2024
  • // CUDA Toolkit Link! ► developer.nvidia.com/cuda-dow...
    // Join the Community Discord! ► / discord
    The graphics card is arguably the most common centerpiece of a PC build. However, hoes does one actually use the GPU, and how can you write code to access the specialized hardware inside these coprocessors? This video explores the concepts needed to make your first CUDA program, and explores the syntax as well as what it all means! Strap in, because it's a lot to digest!
    Test System Specs:
    - Intel Core i7-11700k
    - Gigabyte Z590 UD motherboard
    - 64GB 3200 MT/s Silicon Power DDR4
    - Intel Arc A750 Limited Edition
    - WDB SN770 1TB PCIe 4.0 for games
    - Gigabyte RTX 3070 ti Gaming OC
    - Samsung 970 Evo 500GB PCIe 3.0 for OS/Boot
    - 650W EVGA PSU
    A GPU is a separate bank of ALUs and FPUs, designed to offload heavily parallelizable calculations from the main CPU onto a dedicated chip. The Graphics Card refers to the entire sub-system, containing the PCB, Memory, and Cooling solution.
    Building a Budget PC can be tough. Not only are GPUs and CPUs so incredibly expensive, but they can be hard to find on a budget... But, there are tips and tricks to finding you your dream Budget GPU, and pairing it with a CPU that will give you the performance you want!
    Also, if you're reading this far - I've got more content coming!
    Have a Great Day!
    - Proceu
    Timestamps:
    0:00-0:38 Intro
    0:39-1:18 Preface
    1:19-2:25 Parallelization
    2:26-5:33 Types of Parallelization
    5:34-5:55 Other GPU Hardware
    5:56-6:26 Getting Set Up
    6:27-6:40 Default File
    6:41-8:00 CUDA Headers
    8:01-9:18 Kernel Property 1
    9:19-9:51 Kernel Property 2
    9:52-10:24 Kernel Property 3
    10:25-11:24 cudaMalloc
    11:25-13:15 cudaMemcpy
    13:16-15:32 Writing GPU Code
    15:33-16:04 cudaDeviceSynchronize
    16:05-16:16 Please Free Your Variables!
    16:17-16:37 cudaSetDevice
    16:38-17:35 Test Out Your Program
    17:36-19:15 Conclusion
    19:16-19:36 Guinea Pig Cam
    #nvidia #rtx #cuda #budgetgpu
  • НаукаНаука

Комментарии • 3