As expected, NVIDIA today unveiled an updated family of mobile graphics cards based on Turing GPUs. It registered several new video cards of the GeForce RTX Super and GeForce GTX series, as well as some of the previous models that have changed their positioning.
The flagship of the updated family was the video card GeForce RTX 2080 Super. It is built on the same Turing TU104 GPU with 3072 CUDA cores as the desktop RTX 2080 Super. Only the clock frequency differs, which for the novelty in Boost mode reaches from 1080 to 1560 MHz. The frequency depends on power consumption, which ranges from 80 watts in the Max-Q version to more than 150 watts in the regular version. Also new is equipped with 8 GB GDDR6 with a frequency of 1750 MHz, while the desktop GeForce RTX 2080 Super has a faster memory.
Finally, the third novelty was the GeForce GTX 1650 Ti, which in general did not go very far from the usual mobile GeForce GTX 1650. The novelty will be able to offer a graphics processor with 1024 CUDA cores and clock speeds in Boost mode from 1200 to 1485 MHz. There is also 4 GB of GDDR6 memory with a frequency of 1500 MHz. The power consumption level is from 35 to 55 watts.
As for positioning, the GeForce RTX 2080 Super and RTX 2070 Super video cards replaced the regular GeForce RTX 2080 and RTX 2070, respectively. The regular GeForce RTX 2070 was a replacement for the RTX 2060, which in turn took the place of the GeForce GTX 1660 Ti. The existing mobile GeForce GTX 1660 Ti and the new GTX 1650 Ti will replace the GeForce GTX 1650, which in turn is now positioned as a replacement for the GeForce GTX 1050. And if it’s simpler: laptops with GeForce RTX 2060 graphics cards will now cost from $ 999, and with GeForce GTX 1650 – from $ 699.
Along with the new graphics cards, NVIDIA introduced several new technologies inherent in low-power graphics cards (Max-Q). The key one is the Max-Q Dynamic Boost technology, which allows you to adjust the power, and accordingly the frequencies of the GPU and CPU together. That is, in some cases, it allows you to limit the power of the CPU to give more power to the GPU.
AMD SmartShift technology offers something similar, but there we are talking about CPU and GPU links from AMD itself, while NVIDIA seems to have had to work with Intel. Therefore, it will be interesting to see how this will be implemented in practice. So far, NVIDIA promises a performance increase of less than 10%.