Nvidia might announce new GPUs at CES, a few of which we already find out about, however in response to a press launch from Nvidia itself, the corporate plans to take the wraps off of a number of reasonably priced cellular choices, together with the MX550, MX570 and one which caught us unexpectedly, the RTX 2050. These GPUs are slated to seem in laptops to be launched someday within the Spring of 2022 (Autumn 2022 for the southern hemisphere of us).
An RTX 2050 Turing technology GPU we hear you ask? No, it isn’t. The RTX 2050 isn’t a Turing half as we might count on, however an Ampere technology product, based mostly on the GA107 GPU. This is similar GPU that’s slated to seem in upcoming RTX 3050 playing cards. The chip crunch is resulting in some relatively odd merchandise that in any other case wouldn’t see the sunshine of day.
Usually, we’d criticize Nvidia for renaming a 3 yr outdated product and re-releasing it, however that’s not what Nvidia is doing right here. It’s really doing the reverse, taking a more moderen technology Ampere GPU and slapping on RTX 20 sequence branding. It’s a relatively odd transfer. Why Nvidia isn’t calling it the RTX 3040 is puzzling.
In accordance with specs obtained by Anandtech, the RTX 2050 will characteristic 2048 cores and a lift clock of as much as 1477 MHz. The weak half is its reminiscence configuration. 4GB of 14 Gbps GDDR6 over a 64-bit bus is unquestionably on the weak aspect. This spec factors in direction of the 2050 being a considerably lower down model of the GA107 GPU. Provided that the 2050 carries the RTX branding, customers will count on some type of ray tracing efficiency, however even with DLSS, its ray tracing capabilities are more likely to be poor. At the very least the TDP quantity appears good at 30-45W. This implies it received’t require a strong cooling resolution.
By the point Spring rolls round, we’ll have subsequent technology built-in graphics to match to the RTX 2050. Placing apart manufacturing agreements and the advertising and marketing worth of that includes RTX branding, Intel’s twelfth Gen cellular CPUs with Iris Xe graphics and AMD’s Rembrandt APUs may go away entry stage discrete GPUs wanting fairly common. We’ll wait and see how these merchandise carry out in the true world earlier than passing judgment.