Synthetic and real-world file transfer and throughput tests are one thing, but the whole point of the Killer NIC is to increase gaming performance, no? To that end, we kicked up some favorites we’ve found are good games to show performance with. Unfortunately, no one at ThinkComputers plays an MMO, so we weren’t able to test against the network traffic of a less time-sensitive game.
In Team Fortress 2, the Killer laptop had pings of 32 – 35 ms against my preferred server, and saw ~157-162 FPS. The Intel laptop saw 29 – 50 ms, ~110-210 FPS. However, the Intel laptop’s performance stayed around 40 ms and 140 fps. The Killer wins here, by as much as 20% for ping and 15% for framerate.
Unreal Tournament 3 has historically not benefitted much from the Killer NIC, and the Killer 1102 is no exception. The Killer laptop saw 190-193 ms pings with ~60 fps, while the Intel laptop saw 198-202 ms pings and a similar framerate to that of the Killer laptop.
Here, you can see me being super serious while setting up my gaming tests. Or not.
As for LAN party performance, I took them to Lanwar’s MillionManLan 10 and Pittsburgh LAN Coalition‘s RetroLAN. My tests there were by no means scientific, merely observational and “what feels right”. I cannot say that there was a marked difference when playing true LAN games such as Team Fortress 2, Call of Duty 4, and Left 4 Dead 2. However, this is do know: as the Killer NIC is meant primarily to aide in reducing the latency of high latency connections to the Internet, games which rely on Internet-based servers for “local” multiplayer will benefit greatly from the Killer NIC, especially at LAN parties where bandwidth is limited.
I have no evidence to back it up, but its very well possible that the Killer NIC could be a competitive advantage in a LAN party environment. There is a caveat here, though. When you’re at a LAN party, you’re connected via Ethernet. No gamer in his right mind is going to chance a wireless connection at a LAN party.
To that end, I reach the first criticism I have of the Killer Wireless-N, and it doesn’t really have anything to do with the card itself. I think it follows that if a laptop has a Killer Wireless-N wireless card, it should also have a Killer E2100 for its onboard Ethernet NIC. This would present a consistent experience for the gamer and hopefully reduce some driver bloat, and maybe some production costs.
Colin Dean has been a writer for ThinkComputers since 2006.