A little more than a year ago, Nvidia launched the graphics architecture Turing, which debuted through a powerful trio of graphics cards with the GeForce RTX 2080 Ti at the forefront. The prefix “RTX” is a collective name for several new technologies that come with Turing, where the parade number among these is hardware accelerated ray tracing.
► Read about the Turing architecture and RTX features
In 2019, Nvidia expanded the Geforce RTX family with the model RTX 2060, which is the least powerful variant in the series. In addition, the company has released new Turing-based cards in the lower price and performance classes in the form of the Geforce GTX 1600 series, where the decline of “GTX” is rumored to be absent ray tracing-support. It is now reported that the 1600 series will be expanded with the GTX 1650 Ti in the future.
The information comes from Chinese Fashaoyou, who writes that players in electronics manufacturing have revealed the Geforce GTX 1650 Ti, which in the near future will fill the gap between GTX 1650 and big brother GTX 1660. According to Wccftech, information about a GTX 1650 Ti has circulated earlier, but recently has been quiet about the matter.
At present, the Geforce GTX 1650 is the smallest card in the GTX 1600 series and has a scaled-down variant of the TU117 graphics circuit under the hood. The circuit is equipped with 896 CUDA cores, 32 raster units and 56 texture units and is accompanied by 4 GB of GDDR5 memory. This can be compared to the big brother GTX 1660, which comes with 6 GB of memory, 1,408 CUDA cores, 88 texture units and 48 raster ditons.
The performance life between the two models is relatively large and the same applies to the price, where the reference level for the Geforce GTX 1660 is 70 USD higher. Thus, there is room for another model both from a price and performance point of view, and Fashaoyou believes that a full-scale variant of the TU117 circuit applies to the GTX 1650 Ti.
Exact specifications are currently shrouded in obscurity, but it is likely to offer 1,024 CUDA cores, 64 texture units and 32 residual units, as well as a retained 4 GB of memory. It is likely that the TDP value will also remain at 75 watts, which means that extra power supply can be skipped.
It is worth noting, however, that few partner variants of the Geforce GTX 1650 stick to the reference level, but instead the manufacturers focus on offering higher clock frequencies and performance. Then an extra 6-pin PCI Express connector becomes necessary, and similar priorities are likely to be common even with partner-made GTX 1650 Ti.
Fashaoyou writes that the upcoming entry-level card will cost around 1,100 Chinese yuan, which is equivalent to about 150 USD – the reference level for the Geforce GTX 1650. Whether this is true remains to be seen, but if so, the GTX 1650 is likely to be reduced in price.
It is more likely, however, that the “Ti” version will have a price tag of around 180 USD, ie between today’s prices for GTX 1650 and GTX 1660. According to information, the new cards will roll out sometime in early October, but Nvidia has so far not presented any official details.