First RTX 2050 and MX550 Benchmarks Have Surfaced the Web

The very first RTX 2050 and MX550 benchmarks have just leaked online and, needless to say, they leave a bit to be desired. They’re not bad per se, but they’re not going to blow anyone’s mind which, in all fairness, was to be expected. These GPUs were never designed to run the latest and greatest AAA titles at a million frames per second.

Instead, they’re geared towards the budget-minded gamer and those who are perfectly content with limiting themselves to Medium(ish) settings. So, if you fall under either of these two categories, you definitely have something to look forward to!

Let’s Talk Numbers

Based on early 3DMark TimeSpy results, both of these GPUs will be significantly less powerful than NVIDIA’s most recent entry-level offering, the RTX 3050. This, frankly, should come as no surprise, especially given their 64-bit memory interface (and a host of other limiting factors). The MX550 will be NVIDIA’s most affordable mobile GPU whereas the RTX 2050 is basically designed to offer “team green’s” most alluring features like DLSS and ray tracing support to as wide an audience as possible.

The MX550 has scored 7,888 and 2,510 points in the CPU and GPU tests, respectively. The RTX 2050 is a fair bit more impressive with 7,779 and 3,369 although these are by no means earth-shattering results. It’s also worth noting that we don’t know which CPUs these graphics cards were paired with nor the overall testing methodology.

Plus, as if that wasn’t enough, NVIDIA is surely still tweaking its drivers which means that these numbers could, by all means, improve once these graphics cards hit the market.

Middling At Best

For comparison, the RTX 3050 scores around 4,859 points (GPU) whereas the GTX 1650 holds its own with a respectable 3,634. The RTX 2050 is therefore a GTX 1650 with DLSS and ray tracing support. If that sounds like something you’d be interested in, then great! For most gamers, however, that’s probably a bit too underwhelming.

The MX570, a GPU whose power still hasn’t been measured, will be based on the same GA107 Ampere die but will have just 2GB of GDDR6 VRAM so don’t expect any miracles. The same goes for the 2050’s ray tracing capabilities — it won’t be a feature you’ll ever use much because of its meager RT core count and limiting memory bus. Still, it’s a “nice-to-have” and it might be put to good use, depending on the title and overall context.

DLSS, on the other hand, will give the RTX 2050 a fighting chance when it comes to today’s most popular titles, but it would still be wise to temper the hype until we see some real-world benchmarks.

In any case, we’ll get a ton of information on January 4th, so stay tuned!