hinch said: Well Ada is essentially rebuffed Ampere with a new node, enhancements and a crap load of cache added. Blackwell is going to be a lot different as architecture goes (supposedly) so I think they'd keep power consumption to a reasonable level. As in I can't see them going over 500W for the 5090. |
I get what you're saying but Turing, Ampere, RDNA3 were all neutral or underperformed node gains. The last time a new architecture had a meaningful jump in efficiency with other factors being equal was with Maxwell and RDNA1. That would mean simply refined architectures have been faring just as well in efficiency, on average, so I'm not sure there's a trend here.
That being said, at least for the next generation, there's room between TDP and non-RT gaming in the 4090, as well as the possibitlity of making a bigger chip with safer clocks even with meager efficiency gains in 3nm...