Kopite7kimi, a renowned leaker has revealed the specifications of the forthcoming NVIDIA GeForce RTX 3050. The leaker has a long history of revealing important insights about NVIDIA products.
If we go into details about NVIDIA GeForce RTX 3050, it is based on the Ampere architecture. What’s more, the chipset has every feature to go toe-to-toe against top graphics cards.
Also Read: Xiaomi Upcoming High-silicon Lithium Battery to Offer 10% More Capacity.
As per the rumors, NVIDIA initially planned to pack the RTX 3050 with the GA107. However, recent reports suggest that the manufacturer has dropped that idea, and now it will feature GA106 silicon instead. There can be multiple reasons for that. One such reason is GA106 is already there on GeForce RTX 3060. And, NVIDIA might be able to recover any dies that don’t match with the specs of the GeForce RTX 3060. The company can later use them on RTX 3050.
Moreover, the GeForce will arrive in two variants; they both differ in the number of CUDA cores and memory capacity. The GA106-150 die is specially made for the GeForce RTX 3050 8GB. On the other hand, the GA160-140 die will arrive on the GeForce RTX 3050 4GB GPU. So, it is clear that the 8GB model has better stamina than the other two.
Furthermore, the GA106 chipset has a total surface area of 275mm2. It can accommodate up to 30 streaming multiprocessors. In other words, the chipset can feature up to 3,840 CUDA cores. The GeForce RTX 3050 8GB and 4GB could feature up to 20 and 18 SMs, respectively.
Also Read: Google Releasing Firmware Update for Pixel Buds A-Series.