Asked about What the hardware is doing
Why are GPUs used for deep learning?
An answer that works
The core operation is matrix multiplication, which is thousands of independent multiply-accumulate operations. A GPU has thousands of simple cores suited to exactly that, while a CPU has a few complex ones.
Reading an answer is not being able to give one. The section that teaches this is What the hardware is doing, 5 hours of GPUs & accelerators.
Learn this properlyAlso asked about GPUs & accelerators
What is mixed precision training?Your GPU shows 30 percent utilisation. What is likely wrong?Is LLM inference compute-bound or memory-bound?What does kernel fusion do?How do you tell whether a GPU workload is limited by compute or memory bandwidth?You need to serve a 7B model cheaply at moderate traffic. How do you choose hardware?