The Chinese company Galax was the first manufacturer of graphics cards to announce the modification of the GeForce GTX 1650 based on the 12-nm core Nvidia TU106. This GPU, we recall, was previously used in the graphics accelerators GeForce RTX 2060, RTX 2060 Super and RTX 2070. The new device is called the GeForce GTX 1650 Ultra.
Of the physically present 2304 CUDA cores in the TU106 chip, less than half are active, namely 896 units. “Under the knife” also got tensor and RT-cores, responsible for hardware acceleration of ray tracing. GPU frequency is from 1410 to 1590 MHz in dynamic overclocking. The video buffer is represented by four gigabytes of GDDR6 memory with a 128-bit bus and an effective frequency of 12 GHz.
Perhaps the main difference between the GeForce GTX 1650 Ultra and the already familiar model with GDDR6 memory on the Nvidia TU117 chip lies in the TDP level that has grown from 75 to 90 watts. To connect an additional power source at the end, a 6-pin connector is available. The video card cooling system consists of an aluminum radiator and a pair of ~ 80 mm fans.
The cost of the Galax GeForce GTX 1650 Ultra has not been announced. By the way, today the GeForce GTX 1650 series of video cards includes four modifications.