Nvidia going crazy with two more GeForce GTX 1650 variants
by Rustam Iqbal
Two unreleased variants of the GeForce GTX 1650 graphics card have just been added by AIDA64, it has increased the total number of GeForce GTX 1650 GPU’s different variants to four. This time around, Nvidia has outdone themselves, even by the chipmaker’s own standards.
Debuted with the Turing architecture TU117 silicon, the original GeForce GTX 1650 was released last year. The GeForce GTX 1650 graphics card have 896 CUDA cores and are paired with 4GB of 8Gbps GDDR5 memory across the GPU’s 128-bit memory interface.
Nvidia rolled out the GeForce GTX 1650 GDDR6 variant around a little less than a year later, this GeForce GTX 1650 variant cam with the faster 12Gbps GDDR6 memory. However, the GDDR6 variant of the GeForce GTX 1650 had lower clock speeds and the upgrade to the GDDR6 memory allowed the GeForce GTX 1650 GPU to deliver up to 50% higher memory bandwidth. This upgrade to the GeForce GTX 1650 GPU did actually help in the real-world improvement in the performance of up to 14% over the vanilla version. But, now the AIDA64’s latest changelog does show that Nvidia is not done with their GeForce GTX 1650 just yet.
|GeForce GTX 1650 TU106 **||GeForce GTX 1650 TU116 **||GeForce GTX 1650 GDDR6||GeForce GTX 1650|
|GPU Architecture||Turing – TU106||Turing – TU116||Turing – TU117||Turing – TU117|
|Base Clock Rate||1410 MHz||1410 MHz||1410 MHz||1485 MHz|
|Boost Clock Rate||1590 MHz||1590 MHz||1590 MHz||1665 MHz|
|Memory Clock||12 Gbps||12 Gbps||12 Gbps||8 Gbps|
|Memory Capacity||4GB GDDR6||4GB GDDR6||4GB GDDR6||4GB GDDR5|
|Memory Bandwidth||192 GBps||192 GBps||192 GBps||128 GBps|
|Transistor Count||10.8 billion||6.6 billion||4.7 billion||4.7 billion|
** Specifications are not confirmed.
The two new variants of the GeForce GTX 1650 GPU are being referred as the GeForce GTX 1650 TU106 and the GeForce GTX 1650 TU116 by AIDA64. It has pretty much the same ingredients as the GDDR6 version that denotes the type of memory in the model name. The suffixes designate the silicon that is used inside each of the refreshed model this time.
The TU116 die, in case of the GeForce GTX 1650 TU116, is the same one that Nvidia uses in the other GeForce GTX 16-series SKUs, that includes the GTX 1650 Super, GTX 1660, GTX 1660 Super and GTX 1660 Ti.
On the other hand, we have the TU106 die; this is a very interesting selection and the one that actually raises quite a lot of questions. For context, the RTX 2060, RTX 2060 Super and RTX 2070 have the TU106 present inside these Nvidia’s GeForce RTX 20-series models. The TU106 does have the Nvidia’s Tensor and the RT cores, unlike the TU116 and the TU117 silicons. The main question that rises now is that, have Nvidia finally caved in and have decided to bring ray tracing to a GTX 16-series GPU? For the time being, we don’t know that yet, unfortunately. There were no hints given by AIDA64 on the specifications. But, the base specifications for both of the new GeForce GTX 1650 variants should be the same.
Nvidia coming up with the new variants of the GeForce GTX 1650 GPU is probably not just a coincidence. It is likely that Nvidia is just reprocessing the defective dies that don’t meet the requirements for the higher tier models and just using them in the GeForce GTX 1650 variants. Another reason may be that, the chipmaker might just be getting rid of the remaining Turing silicon and trying to make more space for Ampere, which is heavily rumored that it will debut in September.
The Nvidia GeForce GTX 1650 would probably be the best candidate to use and get rid of the unused silicon. That is because, according to the Steam Hardware Survey for May 2020, the GeForce GTX 1650 was the fifth most popular GPU among the Steam gamers. So, if Nvidia needed a fast way to get rid the unused silicon, the GeForce GTX 1650 would be the best candidate.