NVIDIA adds GeForce GTX 1660 to Turing-based GTX lineup — up to 15% faster than the GeForce GTX 1060 6GB
The much-rumored GeForce GTX 1660 is finally official. This GPU enters as an even more affordable Turing-based offering than even the NVIDIA GeForce GTX 1660 Ti. It comes with all the benefits of the Turing architecture, sans the RT and Tensor cores featured in the GeForce RTX cards.
Thanks to these new advances brought by the new Turing shaders, NVIDIA claims the GeForce GTX 1660 performs 15% faster than a GeForce GTX 1060 6GB and up to 30% faster than a GeForce GTX 1060 3GB on average. All while sporting less CUDA cores and a similar 120W TDP.
NVIDIA also went on to share that the GeForce GTX 1660 is on average 68% faster than a GeForce GTX 970 and more than twice as fast as a GeForce GTX 960.
The following are the specifications of the GeForce GTX 1660:
NVIDIA GeForce GTX 1660 Specifications
- TU116 @ up to 1785 MHz, 12nm FinFET
- 1408 CUDA cores
- 6GB GDDR5 @ 8 Gbps, 192-bit memory interface
- 120W TDP
- 1x 8-pin PCIe power connector
- From $219 (~RM896)
As you can see the NVIDIA GeForce GTX 1660 is a cut-down GeForce GTX 1660 Ti. It also drops GDDR6 in favor of more affordable GDDR5 memory, and comes with the same 120W TDP. It would be interesting to see if the GeForce GTX 1660 can be overclocked to nip at the heels of the GeForce GTX 1660 Ti.
There will be no Founders Edition for the NVIDIA GeForce GTX 1660, and all GeForce GTX 1660 cards will be offered by NVIDIA’s AIB partners.
Pokdepinion: Perhaps a great upgrade for anyone coming from an older GPU…