NVIDIA's Hungry GeForce RTX 4080 Power Draw Is Allegedly 100W Higher Than RTX 3080

It's not really news to tell you that the next generation of graphics cards is going to be power-hungry. Truth be told, we don't have any official information on that point whatsoever, neither from AMD nor NVIDIA. The writing is on the wall, though; besides the constant leaks implying that the next GPU releases from teams both red and green will be monstrous, power-thirsty beasts, there's also the creation of that 12VHPWR connector.


Well, yesterday, kopite7kimi revised his earlier assessment of the proposed GeForce RTX 4080's power demands, stating that the "possible RTX 4080, PG139-SKU360, has a 420W TGP". He also noted that this product would be based on an AD103 processor. The only "103"-series chip that NVIDIA has ever created is the rarely-seen AD103 that is known to be used in some laptop GeForce RTX 3080 Ti models, so that's pretty interesting, too.

Keep in mind that these cards are once again going to be using chips fabricated at TSMC, and on a smaller process than the Ampere GPUs. The GeForce 3000 series is fabricated on Samsung's 8N process, which is thought to be slightly inferior to TSMC's N7. The Ada Lovelace family is going to be fabricated on TSMC's N5, and yet the power budgets are still going up, tier-for-tier.

So saying, we expect some seriously stupendous performance out of NVIDIA's new GPUs when they arrive. If earlier leaks are accurate, that should happen starting in August.