The NVIDIA GeForce RTX 4070 graphics card is now reported to feature 7,680 cores

NVIDIA GeForce RTX 4070 gets a bump in the specs: 7680 cores, 12GB GDDR6X memory, 300W TB, matching the RTX 3090 Ti

The same leaker had previously posted the same configuration for the RTX 4070 Ti, but it looks like the NVIDIA GeForce RTX 4070 will replace it after the alleged spec bump. The card now has the same amount of cores and memory albeit with a much lower TGP rate than previously reported.

The NVIDIA GeForce RTX 4070 is expected to use a full-fat AD104 GPU core with 7,680 cores, or 60 SM cores. From previous leaks, we also know that the AD104 GPU will come packed with 48MB of L2 cache and up to 160 ROP which is simply insane. That's a 25% increase in core count and a 12x cache increase versus the GeForce RTX 3070 Ti rocking two GA104 GPU cores.

Clock speeds have yet to be confirmed, but given that the TSMC 4N process is in use, we would expect clocks in the 2.0-3.0GHz range. The higher-than-normal clock speed comes from the fact that NVIDIA is making a two-node jump considering that the Ampere GPUs with Samsung's 8nm node were actually a 10nm process node with some improvements. NVIDIA skips 7nm and goes straight to the 5nm node not even the vanilla variant but an enhanced version of it. With Pascal on the TSMC 16nm node, NVIDIA delivered a massive jump in frequency and we can expect a similar jump this time around as well.

In terms of memory specs, the GeForce RTX 4070 is expected to outperform the 12GB GDDR6X which is said to be clocked at 21Gbps via the 192-bit bus interface for 504GB/s bandwidth. While the RTX 4070 Ti has been reported to get 400W TBP, the RTX 4070 will get 100W TBP under 300W. Now, this may be the power limit for custom builds with the Founders Edition roaming around the 255-285 power limit. The graphics card is expected to be based on the PG141-SKU331 PCB design. The graphics card is expected to score 11,000 3DMark Time Spy points which will easily match the RTX 3090 Ti graphics card.