Not less than 400 watts, a figure that many fear, since it would represent a power supply of a quantity of watts and a quality that they cannot afford. This is the scenario that we are going to see in less than a year according to the rumors and as such, we must investigate all the reasons that we know today to try to put white on black. Why is consumption not reduced if a new node opens?
TSMC node at 5nm not as good as they paint it, Intel has a chance
The truth is that AMD and NVIDIA are going to bet everything on the new 5 nm of the Taiwanese due to the improvements that this new invoice lithographic process brings, already with EUV under its belt. If logic prevailed that consumptions would return to normal at 250-280 watts per card, reality will show otherwise.
There are two main reasons that explain both this and what is to come, at least to a greater or lesser extent in the coming years. Both Intel and TSMC and Samsung are focused on specific areas of the nodes due to the move from FinFET to GAA, which is posing a challenge and therefore, certain improvements are needed in fields that do not prioritize efficiency.
GPU consumption at 5 nm, better node is not synonymous with better consumption
To be specific, the Big Three are battling two problems and flaws:
- The new lithographic processes are not oriented to achieve a reduction in consumption or simply maintain consumption at the same gain as the space that is achieved. The density achieved by the reduction of the transistors is not equivalent or equitable with the consumption per mm2. The problem here is that both AMD and NVIDIA will need to manufacture chips without the so-called “dark silicones”, that is, without parts of the silicon that are not usable, so they need all the mm2 to the edge of each die to introduce the largest number of transistors.
- And here comes problem number two: the competition between AMD and NVIDIA. Unlike other generations, AMD is now a rival to NVIDIA, so neither of the two is going to keep anything up its sleeve and they are going to put everything available in their new architectures, up to the limit. Such an abrupt hardware jump (+ -70 / 80% more transistors) is a challenge in terms of consumption, because the frequencies do have to be maintained or even go up.What is achieved here is the best that both companies can technically offer, which added to point one results in a beastly increase in performance compared to previous years, but also higher energy consumption.
Therefore, both Navi 31 and AD102 are going to change the paradigm again and the global GPU market, the problem for the user (price and availability aside) is going to be the PSU requirements for some monsters that are going to play. ceiling at 480 watts. Will we need sources of 1000 or 1200 watts to keep them at bay? Will the cards once again suffer problems due to the fast and aggressive peaks due to not having PSUs ready for it?