Truncation is an approximate value method, which means that when a number needs to be rounded to a certain bit, all the values after that bit are discarded, except the values before that bit. This method is often used when the numerical accuracy is not high or a large amount of data needs to be processed quickly. Truncation can process data quickly, but truncated data may have some influence on the final result.