Nvidia’s forthcoming RTX 2050 laptop GPUs don’t seem particularly “new,” given that the RTX 3050 and 3050 Ti GPUs for laptops were announced earlier this year (via VideoCardz). However, it’s apparent that Nvidia isn’t ready to retire its 20 series GPUs just yet; the RTX 2050 will be available in laptops in the spring of 2022, according to Nvidia.
The RTX 2050 is more than simply a resurrected piece of old technology. The RTX 2050, unlike previous 20-series CPUs, is built on the Ampere GA107 architecture rather than Turing.
Ampere is the same architecture that powers the RTX 3050 (and all other 30 series cards), thus putting it inside a 20-series card might potentially deliver some of the same ray-tracing benefits at a lower cost. The RTX 2050 contains 2,048 CUDA cores and 4GB of GDDR6 memory, much like the RTX 3050. However, compared to the RTX 3050’s 128-bit memory bus, it has a 64-bit memory bus, which might result in a performance hit.
While the move might be a means for Nvidia to give a middle ground between the RTX 3050 and the RTX 1650 owing to the continuing chip scarcity, it could also be a way for Nvidia to deliver a GPU that isn’t part of the latest and best 30 series. Nvidia may desire a mid-range GPU that outperforms the RTX 1650 while remaining more inexpensive than devices with the RTX 3050.
Nvidia also launched the entry-level MX550 and MX570 laptop CPUs with the RTX 2050, although the firm hasn’t released much specifics about their specifications. Nvidia just cites “more CUDA Cores” and “higher memory speeds” in passing, with no concrete figures to back it up. All of this comes only two weeks after Nvidia announced a new version of the RTX 2060 — which was first released in 2019 — with 12GB of video RAM instead of the ordinary model’s 6GB.