NVIDIA GeForce RTX 4060 GPU might actually be slower than RTX 3070 Ti

Kopite7kimi recently tweeted that NVIDIA’s GeForce RTX 4060 graphics cards would not be particularly quick and might use an x8 PCI interface. The NVIDIA AD106 GPU, which is supposed to power the RTX 4060 series graphics cards, has a poor performance, according to the leaker.

Although we are unsure of the precise configuration being discussed, it is more likely that he is referring to the full-fat die than a scaled-down version. According to a leaked 3DMark Time Spy Extreme benchmark, the GPU scored less than 7000 points.

With this performance, the AD106 GPU should be slower than the RTX 3070 Ti and comparable to the RTX 3070, which is not the significant improvement that the higher-end AD102 and AD103 GPUs are receiving.

The AD106 GPU (GeForce RTX 4060) should be up to 18% quicker than the RTX 3060 Ti and 69% faster than the RTX 3060 with a Time Spy Extreme score between 6500 and 7000.

Even while a performance improvement of more than 50% is significant, the 2x leaps that have been rumoured for high-end cards are larger. Gaming would at most see a 30–40% improvement if synthetic benchmarks increased by 50–60%.

The leaker further states that the PCIe x8 interface will be used by the “Ada Lovelace” GPUs AD106 and AD107. It makes sense given that these GPUs are primarily aimed at the mobility market (Laptops/Notebooks). The GeForce RTX 4060 and 4050 for desktops will now have an x8 link, which shouldn’t be a problem for most PCs as next-generation systems from AMD and Intel support Gen 5 PCIe slots. However, using an older Gen 3 platform can cause problems due to its lower bandwidth and transfer speeds.

Also Read:

• NVIDIA Hopper H100 GPU broke every record on MLPerf AI Benchmark

Source