As NVIDIA’s GeForce GTX 1660 Super inches closer to reveal, the specifications of the GPU have become more or less clear. As per rumors, the GTX 1660 Super will essentially be the vanilla 1660 with faster GDDR6 memory. The rest of the details will be identical and we expect it retail at the same price as well.
The GPU is expected to come with a 192-bit bus, 6GB of GDDR6 memory and 1408 shaders pulsating at a base and boost clock of 1530 and 1785MHz, respectively (the clocks will vary from OEM to OEM). Like the original 16 series cards, there will be no FE cards in case of the Supers either.
Today’s source is a Chinese retailer with details of a regional brand, Maxsun whose variant features a dual-fan heatsink with a moderately fancy design and a 6-pin power connector. The GTX 1660 Super’s only advantage over the vanilla 1660 will be the faster memory which will, in turn, result in higher bandwidth. The memory is expected to be clocked at 14Gbps and when paired with the 192-bit bus should culminate in a total bandwidth of 336GB/s.
GeForce GTX 1650 Super?
Seeing that NVIDIA has created Super variants of pretty much the entire Turing lineup, we expect the 1650 to get one too. That’s mainly because these new cards are a response to AMD’s Navi lineup and as little Navi hits retail, NVIDIA will want to have all the Supers up and running. The GTX 1650 was already more of a disappointment, performing worse than the RX 570, so we can hope for a faster variant to battle the Radeon RX 5500.
Seeing that all the newer Turing cards are getting the Super suffix, I don’t really see the 1650 getting the same treatment, rather than a Ti. For the GTX 1650 Super to be relevant, it needs to be at least 30-40% faster than the 1650 otherwise it’ll be eaten for breakfast by the RX 5500. We’ll keep you informed as we hear more about the GTX 16 series Supers.