Tweet
yes agreed but provided the measurement let you set the interval to measure the number of frames in you could just set it lower to counteract this, and if you were recording minumum frames per 1/4 second its likely that at somepoint if you played for long enough you would catch a period when very few frames were generated, the only reason we measue in frames per second is that it was a logical standard, nvidia exploited and still do exploit this, there is no reason why benchmarks could not move to frames per quarter of a second as measuring the number of frames generated takes next to no overhead at all. there must be a benchmark out there which can be modified to take smaller time periods, i am sure that this would expose nvdias cards for what they are if measured this way. wouldnt that be funny nvidia made to look like idiots by the one thing they treasure, benchmarks
)
Comment