The scenario is between daunting and worrying, maybe both states at the same time, but the market goes towards maximum performance with the best efficiency, but… It is not always like that. The best example we’re going to have in GPUs, since TSMC’s 5nm (as we reported over a year ago) won’t be the panacea that many are predicting and if we add to that l monolithic architecture from NVIDIA. .. give us a RTX 4090 where its consumption reaches 550 watts.
If we look back, we will see that while the power of GPUs has increased a lot, rather a lot, consumption has also increased dramatically. Blame it on two differentiating factors and a situation we haven’t seen in a long time, where AMD and NVIDIA, as well as TSMC and Samsung are to blame, and it’s (the fault) that we’re going to suffer in the RTX. . 4000 and RX 7000.
Competence, process and technology
These are the three factors that explain what in just a year we will have on the table, we will not expand too much because we treat the problems independently, but it is practical to put them together and explain them to understand what NVIDIA will do put on the shelves. The first factor, competition, is obvious: AMD is a serious threat to NVIDIA and with Intel entering the market, you must do everything you can to win the battle.
The second factor, the lithographic processes, more precisely the 5nm of TSMC which will be used by the two companies. Advanced, yes, expensive, of course, optimal, not at all. N5 It’s a step up in terms of efficiency, density, and maybe performance, but if you massage it into large dies what we find is a very expensive piece of silicon that doesn’t make up for the jump made. and that is its main problem. Apple “suffers” with its new SoCs paid for at the price of gold, NVIDIA and to a lesser extent AMD already know what they are exposed to.
The third factor is technological in terms of architectures. AMD is going to bet on the MCM in a much smarter move than NVIDIA, while the latter does not have the architecture that will support multiple ready-made matrices, which explains the reason for this news and which we will talk about next.
NVIDIA RTX 4090: 5 nm and 550 watts of consumption
If the 350 watts of the current RTX 3090 and the 480 watts that they manage to consume with overclock seems barbaric to you, hold on to the chair, because the rumors of the previous 450 watts apparently remain a mere anecdote.
First of all, let’s say you take a little salt for the sequel, as we’re right in the middle of rumors and speculation, so let’s go. Ada Lovelace will have a TGP (eye, not TDP) which must not be less than 550 watts,
Why has this been revealed now? Much has to do with the publication of the 600 watts PCIe 5.0 and the connector 12 M-Molex pins that we saw last week and thus paved the way for this leak. Truth, lie or reality? We know that the dice are going to be gigantic, that they will go to the limit of frequencies and that they will almost double the number of Shaders, all in a monolithic dice and with the competition of Navi 31 blowing from behind and whispering in the ‘hear.
If the leaks are true, AMD will hit more than 150% compared to the current RX 6000, NVIDIA cannot therefore afford what previous generations have achieved: slackening performance and consumption as well as creating economically profitable chips. Now everyone is going to the limit and that leads the knot of yesteryear to a scenario that is not very contemplative in size and efficiency, as well as in temperature. We’ll see … Also go and prepare your pocket for 1500 watt power supply at least for your new PC if you want a TOP GPU.