GeForce MX570 graphics test in OpenCL – as fast as 2050

Considering the GeForce MX570 laptop graphics test in OpenCL, this example does not seem to be a bad choice. The MX570 was introduced last December, but even after 4 months, few details are still available.

A laptop called HP Zhan 66 14-inch G5 on the website Geekbench Observed, after passing the OpenCL test. This graphical API is of course not ideal for testing the performance of graphics, especially when considering gaming, however it is the only test we currently have.

The GeForce MX570 is based on the Ampere GA107 chip but with limited memory. Unlike most RTX cards, this model only uses 2 GB of storage and 64-bit bus. Nvidia initially claimed that the MX570 had limited beam tracking capabilities, but later found that we were actually seeing more limited performance due to low clock speeds. The Green Team does not disable RT or Tensor cores, which is why the MX570 supports Ray Tracing and DLSS.

GeForce MX570 graphics test in OpenCL

Compared to the RTX 2050, the MX570 uses the same graphics chip with the same number of cores (2048). However, its thermal power is 15 to 24 watts, while the RTX model is in the range of 30 to 45 watts. This alone should affect performance. The interesting thing is that according to the new (and previous) revelation, this did not happen.

In the OpenCL test, the GeForce MX570 is as fast as the RTX 2050. Unfortunately, Geekbench did not share graphics power data, which will be very important here. The RTX 2050 can score from 45986 to 48617. The MX570, however, scored 44,683.

Nvidia has previously announced that the MX55 / MX570 laptops will be available this spring.