Results: Battlefield 3
Originally, we weren’t going to use Battlefield 3. It’s on the older side, Battlefield 4 is coming soon, and the in-game sequence we use for benchmarking isn’t particularly interactive. Instead, we planned on testing with Tomb Raider. Unfortunately, Tomb Raider’s in-game save system makes it difficult to play through the same sequence without overwriting auto-saves. Very quickly, we ran out of save spots and realized that falling back to Battlefield 3 was necessary.
With each card hooked up to three monitors, the outcome is almost unanimous. Lacking frame pacing, Nvidia wins five votes to one.
The next set of results comes from six different gamers (giving us 12 unique opinions throughout the day). Attached to one 30” screen at 2560x1600, two participants voted for AMD, two showed favor to Nvidia, and two said the experiences were similar enough to call a tie.
Naturally, it would have been great to have even more opinions, to have the same large group judging the AMD-based platform before and after frame pacing, and to more granularly qualify the background of each respondent to better gauge the audience’s threshold for detecting smooth gameplay. However, doing so would clearly require a longer-term study than one full-day event.
This corresponds really well with what Don found in his story, where Catalyst 13.8 beta makes a big difference to the benchmark numbers in Battlefield.