Saving the Planet with Better AI Data Centers (with Crusoe CEO Chase Lochmiller)
ACQ2 by Acquired
00:00
Latency in Inference and the Impact of Neural Network Execution Time
The latency added by the inference process is negligible compared to the time it takes for large language models or neural networks to process inputs. So, when interacting with an AI application, any sluggishness is primarily due to the execution time of the neural network rather than network infrastructure. These models contain billions of parameters.
Play episode from 11:07
Transcript
