Introduction to the RTX 4070 GDDR6
NVIDIA has recently introduced a new variant of the GeForce RTX 4070, featuring GDDR6 memory at a slower speed of 20 Gbps, in place of the original 21 Gbps GDDR6X. This subtle change has raised questions regarding the actual performance impact of the memory modification. Wccftech has taken the initiative to benchmark a Galax-branded RTX 4070 GDDR6 to shed light on its performance against the GDDR6X variant.
Benchmarking Results Between the Two Variants
Wccftech’s extensive testing involved 27 game tests and 7 synthetic tests, focused on three different resolutions: 1080p, 1440p, and 4K Ultra HD. The core configuration and GPU clock speed of the RTX 4070 GDDR6 remain unchanged, which suggests that the performance difference primarily stems from the GDDR6 memory type and speed.
Performance Insights at Different Resolutions
The results indicate that the RTX 4070 GDDR6 is impressively close in performance, showcasing a delta of 0-1% slower at 1080p and 1440p when compared to the GDDR6X model. However, at 4K resolution, the difference widens to approximately 2%. It is noteworthy that in some specific cases, such as in ‘Metro Exodus’ with the extreme preset, the GDDR6 model shows a 4.2% loss in performance, while ‘Alan Wake 2’ reports a 2.3% decrease. In synthetic tests, the new model performs similarly, exhibiting deltas around 0-0.2% against its predecessor.
