NVIDIA GeForce RTX 2050 Mobile vs AMD Radeon 660M vs NVIDIA GeForce RTX 2080 (Desktop)

NVIDIA GeForce RTX 2050 Mobile

The NVIDIA GeForce RTX 2050 for laptops (GN20-S7) is a fast mid-range gaming graphics card for laptops. The RTX 2050 is based on the modern Ampere GA107 chip (like the RTX 3050) which is manufactured on an 8nm Samsung process (8N). Compared to the similar GeForce MX570, the RTX2050 offers Raytracing and DLSS support.

Compared to the old Geforce RTX 2060 Mobile, the 2050 offers more shaders (2,048) but a greatly reduced memory bus (64 Bit). It is available in different versions with with TGPs ranging from 30 to 45 Watt. The maximum boost clock ranges from 1155 to 1477 MHz. It features up to 4 GB GDDR6 graphics memory (14 Gbps leading to 112 GB/s max throughput).

The gaming performance of the mobile RTX 2050 should be clearly lower than the RTX 2060 thanks to the greatly reduced graphics memory bus. First benchmarks position a fast version slightly above a GeForce GTX 1650 and on the same level of a GeForce MX570. However, games that support DLSS can offer clear performance advantages thanks to the Tensor cores.

AMD Radeon 660M

The AMD Radeon 660M is an integrated GPU in the entry level Ryzen 6000 series for laptops (Rembrandt, e.g. Ryzen 5 6600U). The GPU is based on the RDNA2 architecture and has 6 of the 12 CUs (= 384 shaders) clocked at up to 1,900 MHz (depending on the CPU model). The 680M and 660M iGPUs are the first integrated graphics card with hardware raytracing support, although the performance is too slow to use it in current games.

Due to the halved CU count and lower core clock, the Radeon 660M is clearly slower than the Radeon 680M and most likely only on par with the old Radeon Vega 8 iGPUs. 

Thanks to the modern 6nm process and clever power-saving features, the power consumption is comparatively low (according to AMD), so the graphics card can also be used for slim and light notebooks (depending on the CPU model).

NVIDIA GeForce RTX 2080 (Desktop)