UPONUS,CIFFANDCRYPTOLOGICACOMPARISON.docVIP

UPONUS, CIFF AND CRYPTOLOGICA COMPARISON An UPONUS report prepared by “certain independent third-party testing, as well as internal company testing” is based on a completely wrong measurement methodology, which I will describe in detail below. Let’s start from the first claim: 1. “Virtually no latency in using SASE (encryption technology without compression) to send encrypted files or encrypt streaming data for on-the-fly communication.” Instead of simply measuring actual encryption and decryption time, which can be easily and accurately performed by adding beginning and end time stamps in SASE encryption and decryption source code (which they already changed for the measurement), they first transmit data without an encryption and a decryption between two computers, and then transmit data with an encryption and a decryption between two computers, claiming that the (averaged) difference between the second and the first measurements corresponds to time required to perform an encryption and a decryption. This methodology can be completely wrong in case that software uses buffering and TCP/IP transmission, as in UPONUS test, thus producing a latency even in case without an encryption and a decryption, which after a subtraction from the case with an encryption and a decryption gives wrong result of lower latency encryption and decryption. In other words, with simple programming trick you can get exactly zero time or even negative time as a result, thus claiming that an encryption and a decryption actually accelerates the transmission, which is impossible! To make methodology even worse, “the application was modified to only read one block of data from a file and used that block multiple times.” This approach provides false sense of an encryption and a decryption speed, since data are read from and written into the processor cache, which is not appropriate to real-life scenario. Therefore, all measurements disclosed in tables 1-1, 1-2 and 1-3 are useless since