Big picture: Recent rumors about upcoming graphics cards have raised concerns about rising power consumption. Nvidia and AMD have yet to fully declassify the GPUs they plan to release later this year, but the company revealed some worrying numbers this week in its upcoming roadmap.

- Advertisement -

VentureBeat interviewed AMD SVP Sam Nuffziger this week. included images showing the company’s predictions for near-future hardware. While Naffziger expresses confidence that AMD will be able to achieve its energy efficiency ambitions over the next few years, AMD’s numbers paint a worrying picture of growing energy demand.

- Advertisement -

As engineers everywhere grapple with the limits of Moore’s Law, consumers fear that GPU TDP could rise to alarming levels. Rumor has it that Nvidia’s next-gen flagship, the RTX 4090, which could launch later this year, could need up to 600W. The company’s current top card, the 3090 Ti, consumes 450 watts.

- Advertisement -

If we don’t see 600W cards in the next generation, AMD’s chart suggests we will see them in the next or soon after. It shows a dramatic increase in power consumption around 2018 with GPUs hitting 600W and then 700W until 2025. However, the graph does not indicate whether this only counts top-end hardware. Currently, the most popular cards on the market consume only a fraction of this power, but even future mainstream GPUs will need more power.

Naffziger said that AMD can achieve the performance gains that consumers expect from new graphics cards while addressing the power consumption issue. The company is touting Infinity Cache as a unique selling point, but Naffziger also highlights chipset design as an area in which it leads Nvidia.