Nvidia Corrects Typo, GeForce RTX 4070 Super Confirmed with Full 48MB L2 Cache

Photo of author

By worldnewsdb.com

image: NVIDIA

Nvidia has rectified a typo in its launch materials for the upcoming GeForce RTX 4070 Super, confirming that the highly anticipated graphics card will indeed feature a full 48MB L2 cache, contrary to initial reports of 36MB. The correction sheds light on the impressive capabilities of the AD104 graphics processor, set to power the GeForce RTX 4070 Super.

The preliminary specifications for the GeForce RTX 4070 Super had initially created confusion, suggesting a discrepancy between the stated 48MB L2 cache and the mistakenly mentioned 36MB in Nvidia’s launch materials. The error was promptly addressed and corrected by Nvidia, as reported by Wccftech.

Based on the AD104 graphics processor, the GeForce RTX 4070 Super boasts 7,168 CUDA cores and a 192-bit memory interface. Although it does not utilize the full AD104 GPU found in the GeForce RTX 4070 Ti, featuring 7,680 CUDA cores, the AD104 in the RTX 4070 Super does come equipped with the full 48MB L2 cache, maximizing the benefits of memory bandwidth.

Despite the initial confusion, Nvidia’s website now accurately reflects the specifications, highlighting the GeForce RTX 4070 Super’s 48MB L2 cache and a robust memory subsystem with a peak bandwidth of 504 GB/s.

In terms of compute performance, the GeForce RTX 4070 Super is positioned closely to the GeForce RTX 4070 Ti, boasting 35.5 FP32 TFLOPS compared to the latter’s 40.09 FP32 TFLOPS. With a shared memory subsystem, the performance gap between the two models can potentially be narrowed further through overclocking.

Also Read: “boot loop” bug: why Apple pulling the latest iOS 17.3 beta. whats next?

Nvidia and TSMC’s ongoing refinement of process technologies introduces the possibility of higher overclocking potential for the AD104 GPUs produced for the GeForce RTX 4070 Super. The upcoming release will also see custom graphics cards from Nvidia’s AIB partners, some featuring enhanced TGP (total graphics power) beyond the 220W set by Nvidia. These custom boards are expected to deliver increased overclocking potential, adding an extra layer of excitement for enthusiasts anticipating the GeForce RTX 4070 Super’s performance.

Leave a comment