The transient response performance is disappointing, the 12V rail deviates about 3% when a load of 9.25A is applied at 25% of 50% load. Good units are about in the 0.5% deviation territory. Whenever you game, the power is not constant. For a typical GPU, it fluctuates
heavily. Look at the GTX 1080, which is actually a very
relaxed GPU in terms of power spikes:
From the looks of it, it spikes about 70W the majority of the time up and down. That is 5.8A right there of transients going on constantly when gaming. Although HardOCP tests at 9.25A, per the ATX spec testing methodology, 5.8A will still cause a voltage drop and the Seasonic Prime's performance in this aspect is mediocre at best. Other graphics cards such as the GTX 980 have much heavier spikes, making transient response a more important consideration when purchasing such a card.
I mean, it's still a great unit, but when it comes to the hefty price you bet you'll pay for it, you want it to be great in every aspect, and it is purely mediocre with transient response, which is a disappointment. Transient response is much more important than progressive load testing in my opinion, because once you are in a game, average power does not change
too much but it's the fluctuation that really gets heavy and affects the voltage outputs. This is why I think JG really needs to adopt transient response testing, I'd much rather have that than the less important turn-on transient tests. Just to show you how much these GPUs can spike, look at the R9 290:
It doesn't matter if voltage regulation is top notch because the voltage will be fluctuating a lot from the transients. This is one aspect in which I believe a tier list falls short, some people need a PSU with better transient response performance, and others may have a more relaxed GPU, so in that situation what PSU is good or not really depends on that person and his or her exact needs and can not be categorized into a tier. To put it in InvalidError's own words, transient response testing is definitely one of the two most important things to be tested. I just wish a $200 unit had better performance in it. Nonetheless, the Seasonic Prime units are good, but they are only somewhat better than the Seasonic Platinum units, which were phenomenal in the first place, but these are nothing groundbreaking. The best thing is the efficiency, with Kitguru measuring over 95% at 50% load.