NVIDIA's next-gen GeForce RTX 4090: AD102 could push 100 TFLOPs+

NVIDIA's next-gen Ada Lovelace GPU architecture will be fighting AMD's next-gen RDNA 3 GPU architecture later this year, with both GPUs rumored to provide some rather exciting upgrades in performance over the current generation GPUs.

The new NVIDIA AD102 GPU should be the chip that powers the new flagship GeForce RTX 4090 Ti and GeForce RTX 4090 graphics card -- or, according to some rumors the RTX 5090 Ti and RTX 5090 -- with leaker "kopite7kimi" tweeting that AD102 could have over 100 TFLOPs of compute performance.

100 TFLOPs of power from AD102 would make it 2.5x faster than the RTX 3090 Ti, which spits out 40 TFLOPs and 2.8x better than the RTX 3090 with its 35.5 TFLOPs of FP32 compute performance. Remember: pure TFLOPs doesn't mean = gaming performance, but kopite7kimi does say "Lisa and Jensen's competition will give us a 100 TFLOPs gaming war in a few months" with a key takeaway from that: 100 TFLOPs + gaming war.

• Read more: NVIDIA testing insane next-gen AD102 GPU: freaking 900W power!!!
• Read more: NVIDIA's next-gen GeForce RTX 4090, AD102 GPU: PCIe 4.0, not PCIe 5.0
• Read more: NVIDIA's new AD102-based GeForce RTX 4090: 600W+ power usage CONFIRMED
• Read more: NVIDIA's next-gen Ada Lovelace 'AD102' GPU 'has started testing'

TFLOPs dominance is just one part of this new GPU war: ray-tracing performance, upscaling technology (NVIDIA DLSS vs AMD FSR), display connectivity being bumped to DisplayPort 2.0 (enabling 4K 240Hz and 8K 120Hz), and so much more. AMD is reportedly using an MCD (multi-chip die) design with its Navi 31 GPU while NVIDIA is reportedly going with a monolithic GPU again for its AD102 GPU.

We can expect to see 100 TFLOPs of power from NVIDIA's next-gen AD102 GPU with 18432 CUDA cores at 2.7GHz, which is something we should see: the GeForce RTX 4090 should include a slightly cut-down GPU from the RTX 4090 Ti and other Ada Lovelace-based GPUs, so GPU clocks can be higher.

• Read more: NVIDIA could skip GeForce RTX 40 series, right into RTX 50 series GPUs
• Read more: NVIDIA's next-gen GeForce RTX 4090 rumored to use new GDDR7 memory
• Read more: NVIDIA GeForce RTX 4090 rumored with 'Infinity Cache' style cache
• Read more: NVIDIA's next-gen GeForce RTX 4090: 4K gaming at insane 400FPS+ teased

AMD on the other hand is rumored to use a higher 3GHz+ GPU clock, even with its MCD design on the next-gen Navi 31 GPU and the expected Radeon RX 7900 XT graphics card later this year. Both of the cards will require more power, with NVIDIA expected to use somewhere in the 600W range with its AD102-based GPU... hell, NVIDIA is reportedly testing an AD102 GPU with an insane 900W of power.