The funny thing here is that it could use the same GA107 chip as the supposed RTX 3050 for desktop, where in addition, despite the fact that its name includes it in the 20 series, it will be more powerful than the RTX 2060 Mobile with less consumption, or at least on paper.
And we talk about 2048 shades at a base speed of at least 1155 MHz with a boost of 1477 MHz, so these are numbers that should be left between RTX 2070 and RTX 2060 in laptops. In addition, consumption is extremely low compared to its supposed architectural sisters and this is where we think of an Ampere GPU: 30 watts to 45 watts, still as cTDP which will be resolved by each laptop manufacturer and model.
However, there is no doubt about the memory configurations, since you will have 4 GB of GDDR6 attached to a memory bus of 64 bits
Unless surprise, NVIDIA may have changed their minds about the RTX 3050 for laptops and called it the RTX 2050, so rumors of the former were well underway if that theory is true, which we will never know.
NVIDIA MX570
If the specifications of the RTX 2050 have been lacking, those of its next GPUs are disappointing due to the little and almost no information provided, which makes sense at first to reserve the blow to its rivals, but which reduces its strength in an official version. presentation which hardly brings anything.
What leaked just 4 days ago may well match the RTX 2050 mentioned above, as what NVIDIA provided fits it almost perfectly, but on the other hand the GPU seen was specified with 2 GB of VRAM, therefore, we cannot guarantee anything about this.
The only thing NVIDIA has shown are technical specifications on the technology support, nothing regarding the configuration of the chip or even whether said chip will also be the GA107. Therefore, we know that it will use GDDR6 and support the following list:
- Optimus
- To strenghten
- Soporte para drivers Game Ready
- PCIe 4.0
- DirectX 12
- volcano
- OpenGL
- OpenCL
- Windows 10 and Windows 11
Apart from that, NVIDIA comments in its statement that the MX570 is the fastest GPU in the series to date and at the same time is the first with the 70 designation under its belt. It is specially created for gaming and work thanks to its more power-efficient CUDA cores, while incorporating faster VRAM speed than its predecessors. In addition, NVIDIA leaves the lace to its rivals by saying in its statement that it is the fastest GPU when it comes to photos, video editing and games compared to the latest integrated graphics, which refer to Gen 12. Intel and the supposed Navi who will arrive with Rembrandt.
NVIDIA MX550
It is the direct successor to the MX450 and like the MX570, there is no information on its specifications. As expected, the list of supported technologies is identical to that of its big sister:
- Optimus
- To strenghten
- Soporte para drivers Game Ready
- PCIe 4.0
- DirectX 12
- volcano
- OpenGL
- OpenCL
- Windows 10 and Windows 11
Interestingly and logically, NVIDIA recommends upgrading to an RTX GPU if you want to play faster or are more demanding with gaming or creative performance. However, there is only one date revealed for the launch and it is not precise as such, since NVIDIA cites Spring 2022 as the key.
It also encourages people to wait until the characteristics of notebooks using these three GPUs are revealed, so it is possible and more than likely that all of its partners will stock them since there are hardly any models available on the market at low prices.
It is clear that with these three models and on the aforementioned dates the blow will be against the APU Rembrandt and his iGPU, as well as against the Intel ARC Alchemist Xe LP which will immediately dominate the gaming market after its launch, where NVIDIA would have nothing to compete with if it launched now or after CES 2022. We also do not know exactly where these GPUs will stand in terms of performance. Supposedly the MX570 would be faster than a GTX 1650 SUPER, so the RTX 2050 should be very even with the RTX 2060M, while the MX550 should be able to outperform the normal GTX 1650 without too many issues.
We’ll be paying attention to the specifications of these cards when Nvidia reveals all the remaining information, which is a lot, until then we’ll have to be patient at least until spring.