NVIDIA GeForce RTX 4070 SUPER GPU confirmed to feature full 48MB L2 cache on AD104 chip

NVIDIA has confirmed that the upcoming GeForce RTX 4070 SUPER GPU comes with a full 48MB of L2 cache for the AD104 chipset.

NVIDIA's GeForce RTX 4070 SUPER has 48MB of L2 cache in its AD104 GPU, like the RTX 4070 Ti

When NVIDIA announced the GeForce RTX 4070 SUPER graphics card last week, the spec sheet listed a 20% upgrade to the AD104 GPU core count over the RTX 4070, yet its L2 cache was listed at the same 36MB. Now, NVIDIA has confirmed that the 36MB number was listed in error and has updated its product pages with the correct cache amount of 48MB, which is the same as the RTX 4070 Ti GPU.

Having a higher L2 cache is very important for a GPU like the GeForce RTX 4070 SUPER. It comes with a 192-bit memory bus interface and targets high refresh rate 1080p and 1440p gaming PCs. Additional details like higher resolution textures and the high resolution itself can lead to bottlenecks within the memory subsystem, and while the RTX 4070 SUPER comes with the latest G6X memory, there are still certain scenarios where the memory can become jittered and cause lag in games. Having a large L2 cache near the GPU provides high-speed communication across all GPC devices to send data back and forth.

As such, having 48MB of L2 cache on the NVIDIA GeForce RTX 4070 SUPER GPU is definitely a very good decision and would represent a good improvement over the non-SUPER variant.

NVIDIA GeForce RTX 4070 SUPER GPU specifications: AD104 GPU upgrade with 20% more cores

The NVIDIA GeForce RTX 4070 SUPER will have an AD104-350 or AD103-175 GPU (PG141 SKU 335) with the same 7168 positions in both options. The GPU is expected to get 36MB of L2 cache. The card will stick 12GB of GDDR6X memory via the same 192-bit bus interface. For the RTX 4070 SUPER, the card will have a TBP 20W higher than the current 4070 at 220W.

The card offers 36 Shader TFLOPs, 82 RT TFLOPs, and 568 AI TOPs while supporting all the latest NV encoding technologies such as AV1 and H.264. All of this combined puts the RTX 4070 within 95% of the 4070 Ti's performance.

Superior NVIDIA GeForce RTX 4070 GPU performance: faster than RTX 3090 and only 5% slower than 4070 Ti on average

In terms of gaming performance, the NVIDIA GeForce RTX 4070 SUPER is said to be approximately 15% faster than the RTX 4070 Non-SUPER, representing the biggest improvements in the SUPER lineup. The card is also reported to be faster than the RTX 3090 without framerate although performance will vary on a game-to-game basis. NVIDIA also confirms that the card is only 5% slower than the RTX 4070 Ti, which retails for US$799. This has also been replaced by the RTX 4070 Ti SUPER.

AI is a major talking point for the NVIDIA GeForce RTX 4070 SUPER series and the RTX 40 SUPER series in general, delivering up to 2x faster performance with DLSS 3.5, 70% faster image generation and 50% faster video creation speeds using machine learning improvements through… Tensor. The RT-RT and Tensor-RT LLM stack is available on Windows and is updated on the go.

NVIDIA GeForce RTX 4070 SUPER GPU Power Efficiency: Absolutely Crushed Performance/Watts

Another talking point for NVIDIA GeForce RT 40 SUPER GPUs like the RTX 4070 SUPER is their class-leading performance efficiency. While the GPU is rated at 220W, actual gaming wattage is around 200W on average which is less than half the TGP of the RTX 4090 and well below what the Radeon RX 7900 series consumes. Even the 7800 XT is rated at a much higher 263W TBP . Video playback power is also 16W out of the box while idle power is 11W, something the AMD RDNA 3 GPUs had trouble fixing during launch.

NVIDIA GeForce RTX 4070 SUPER GPU pricing and availability

The NVIDIA GeForce RTX 4070 SUPER GPU will be available on January 17 in both Founders Edition and custom variants, starting at $599. There are plenty of custom templates to choose from like the ones shown below:

NVIDIA GeForce RTX 4070 SUPER Founders Edition Gallery:

NVIDIA GeForce RTX 40 desktop GPU family: