VACnet is a large server farm for detecting traps in online gaming

Valve has become very serious with the users who cheat in CS: GO, the company has introduced VACnet, a system that includes several high-powered servers to detect cheaters, we tell you all the details.

VACnet contains 3456 cores and 8192 GB of RAM to detect cheaters

VACnet is a server farm that includes no less than 1,700 processors, with the ability to add more processors in the future if the processing power requirements so require. This system will be used in CS: GO and other online games, both from Valve and third parties, to detect cheaters.

VACnet consists of 64 Server Blades, each of which includes 57 cores for a total of 3456. 128 GB of RAM has also been added to each Blade, which translates into no less than 8,192 GB of memory. This system has excellent capabilities of artificial intelligence and deep learning, two features that will be used to analyze the repetitions of games in search of traps and tricks…

John McDonald of Valve noted that in 2016 the CS: GO community was continually talking about cheating. That led McDonald to begin investigating Valve’s deep learning, and finally created VACnet, which is currently being used in Overwatch and CS: GO to detect cheaters.

VACnet is designed to send cases to a human moderator for review, with a success/condemnation rate of 80-95%, which shows a high degree of confidence in the system. Claims filed by players have a conviction rate of 15-30%, which makes the VACnet better at detecting cheats than the average player.