The cyber-space of crypto-currency market is a main issue in terms of security and stability. The novelty and the high volatility of crypto-currencies question their speculative nature. Recently, the crypto-currency price exponentially increased and reached an important burst in the end of 2017. The objective of this article is to detect and test the prediction of this crypto-currency market crash using the Log-Periodic Power Law model (LPPL). We consider 2 main crypto-currencies, Bitcoin and Ether. We find that the LPPL model allows to estimate the date of the crash in the crypto-currency market depending on the window sensitivity.