GDDR6. SUPER. Max Q. NVIDIA’s GeForce GTX 1650 family is already laughably complex, but another variant is on the way to ensure continued confusion. As spotted on the company’s official site (via @momomo_us and VideoCardz), GALAX is prepping an “Ultra” version of the GTX 1650, which will slide between the GTX 1650 SUPER and GDDR6 renditions. This variant will utilize the TU106-125 and feature 896 CUDA Cores and a base/boost clock of 1410/1590 MHz. Check out this chart to see how they compare:
| GTX 1650 Variants | GTX 1650 SUPER | GTX 1650 GDDR6 TU106 | GTX 1650 GDDR6 | GTX 1650 |
|---|---|---|---|---|
| GPU | TU116-250 | TU106-125 | TU117-300 | TU117-300 |
| CUDA Cores | 1280 | 896 | 896 | 896 |
| Base Clock | 1530 | 1410 | 1410 | 1410 |
| Boost Clock | 1725 | 1590 | 1590 | 1590 |
| Memory Clock | 12 Gbps | 12 Gbps | 12 Gbps | 8 Gbps |
| Memory | Â 4 GB GDDR6 | Â 4 GB GDDR6 | Â 4 GB GDDR6 | Â 4 GB GDDR5 |
| Memory Bus | 128-bit | 128-bit | 128-bit | 128-bit |
| TDP | 100 W | 90 W | 75 W | 75 W |
What we’re looking at is a SKU that’s essentially identical to the standard GeForce GTX 1650 with GDDR6 memory. The only difference appears to lie in their TDPs: the new TU106-125-based “Ultra” variant is rated at 90 watts, 15 watts higher than the GDDR6 (75 watts).