An unidentified leaker has posted on the Chinese forum Zhihu the possible 3DMark TimeSpy benchmark results with Nvidia's new entry-level RTX 2050 and MX 550 GPUs. Early benchmarks reveal that both GPUs are a significant downgrade compared to Nvidia's latest budget GPU, the RTX 3050, due to the 64-bit memory interface.
A few days ago, Nvidia officially announced three new mobile GPUs for the notebook market, which are the RTX 2050, the MX570 and the MX 550. The new 500 series of the MX line is expected to be a complete update and replacement of the old 400 series, while the RTX 2050 appears to be a new way for Nvidia to offer DLSS and ray tracing support to an even cheaper audience than the RTX 3050.
Entry card benchmarks that are not completely state-of-the-art
The RTX 2050 specs include 2.048 CUDA cores and 4GB of GDDR6 running on a 64-bit bus with only 112GB/s of memory bandwidth. The maximum power consumption is rated for a GSP of 45W with a maximum boost clock of 1477MHz. Nvidia never endowed its cheap Turing GPUs with RT cores and Tensor cores, while the RTX 20-series GPUs were pretty big. To create the new RTX 2050, Nvidia has leveraged its advances for the Ampere GA107 core instead of a disabled TU106 die.
According to the benchmark post, the RTX 2050 and MX570 are basically the same chip. It's unclear if the MX570 will have the exact same specs as the RTX 2050, including its 2048 CUDA cores and GA107 die, or if a few things will be disabled. One of the main differences is that the MX570 only has 2 GB of GDDR6 memory, which would leave it with a very low capacity compared to even graphics cards from several years ago.
The MX550 and MX570, meanwhile, are completely different GPUSs. MX550 appears to have 1024 CUDA cores and to be based on Nvidia's Turing TU116 GPU instead of the Ampere GA107. The MX550 will also support 2 and 4 GB memory capacities with GDDR6 on a 64-bit bus, the same as the 2050 and MX570. Although it may seem that the number of cores is half that of the GA107, keep in mind that Turing has separate floating point and integer pipelines, while Ampere has a dedicated FP pipeline and an FP/INT pipeline. shared. The Ampere architecture is faster, more efficient, and smaller, but the MX550 aims to be a restrained chip for gaming laptops that don't fall into the premium category.
Source: Tom's Hardware
