Has Nvidia been driving the flood of rumors about its next cards for several months? Until now, a certain consensus was found on the leaks with a consequent gain in terms of performance but also an almost delirious flight in terms of consumption. Some new information suggests that this part would be smoke and mirrors. If the latest leak from Kopite is true, the RTX 4080 will have a much more reasonable power consumption than all the assumptions made in the last few months. And obviously, the trend we’re seeing here would cover the entire RTX 4000 line. The RTX 4080 could therefore arrive with a TBP (total card power) of 320 watts, i.e. exactly the same as the current RTX 3080 but with a performance jump of almost 70%.
Wouldn’t Nvidia be the one talking in the leakers’ ears?
For the record, on this generation Nvidia will probably use the 4N process of TSMC (5nm optimized). But how to explain such a reversal on the consumption when several sources have been going in the same direction for weeks? What if these sources were simply from Nvidia? And if Nvidia was not the one who whispers and distills information from the main leakers (who do, in large majority only relay sources and never test the products…). Some will say that we are paranoid.
But there are also other hypotheses that can be elaborated by studying the other “information” that arrives about this RTX 4080. It seems that this card will have an AD103-300 chip. For the first time, an xx80 will not have the high-end version of the family’s GPU. This adds doubts to this information about power consumption. Has Nvidia found the magic formula to consume less or are they simply lowering the specifications of their cards (at least compared to the initial leaks)? One thing is certain: we clearly don’t know anything about these cards and Nvidia is leading us by the nose.
