NVIDIA GeForce RTX 3060 Ti with GDDR6X memory tested – manageable gain

We remember that NVIDIA launched an updated version of the RTX 3060 Ti model in October last year, but we haven’t seen a proper review of this card until now, probably because the availability of these cards was limited. Now the Polish ithardware site has the KFA2/GALAX RTX 3060 Ti G6X OC model available and has done a comparison with RTX 3060 Ti and RTX 3070 cards.

The RTX 3060 Ti features the same GA104 GPU with 4864 CUDA cores as its older sibling. The only real difference is the memory upgrade from 14Gbps GDDR6 memory to 19Gbps GDDR6X version. This means that the maximum theoretical bandwidth has indeed increased from 448 GB/s to 608 GB/s, but at the cost of higher power consumption (200W to 225W).

The card was tested by ITHardware in many games with three resolutions:

The performance increase is of course minimal, about 1 to 2%, but there is one critical thing that should be mentioned first. The ASUS RTX 3060 Ti TUF model used for comparison (based on GDDR6 memory) is a factory overclocked model with its boost increased from 1665 to 1755 MHz, which is a 5.4% increase. Therefore, this is not a very good comparison, as such tests should be performed under similarly specified models. However, this tells us that the new card is faster than the factory overclocked variant. It’s not a double-digit increase, but it’s definitely not 1-2% as the data would suggest at first glance.

In terms of overclocking, the card was successfully boosted to 1815 MHz, which is about 150 MHz higher than the default clock. Beyond that, the memory could still be overclocked to 1263 MHz, which is 20.2 Gbps. These results were possible despite the locked TDP on this model. This brought a further 4 to 6% increase in performance compared to the standard settings.

Source: ithardware