Let’s get started with Nvidia. In this generation of GPUs, the company has pretty much refreshed everything. The RTX 2060, 2070, and 2080 all got the super treatment, so did the 16 series with the 1660 Super and 1650 super, without forgetting the recent update to the GTX 1650 where the base model now comes standard with GDDR6. Well, it looks like its happening again. And unfortunately no, it’s not a 2080ti super. It’s actually another refresh of the GTX 1650. The best part? There are 2 of them.
This information comes from MomomoUS on twitter. In a recent Aida64 Extreme update the software posted compatibility with 2 new GTX 1650, one based on the TU116 chips and another on TU106.
For your information, the original 1650 and the new GDDR6 variants are based on the TU117 Chip. The new 1650 on TU116 would borrow the same chip as a 1650 super and all of the 1660 variants. As for the TU106 based GTX 1650, it shares the same chip as most of the RTX 2060 and 2060 Super. I say most because the RTX 2060 also comes in TU104.
So why would Nvidia do this? Well, compared to the original 1650, both the TU116 and 106 have the new Turing Nvenc encoder as opposed to the Volta one used on the original 1650. This would make these new variants perfect for low budget streaming PC given the new Turing encoder is vastly superior. What is curious is the use of TU106. This chip would still have tensor and ray tracing cores on the silicon. They would just be disabled. It looks like Nvidia is trying to clear out its inventory of defective chips as soon as possible. Maybe the RTX 3000 series is so good that there would be absolutely no point in keeping the 16 series alive. What do you guys think? Let me know down below.