AMD's RX 9050 GPU comes up short in first review, but there's a good reason why it lags behind NVIDIA's RTX 5050 here

Q: How did the RX 9050 compare to the RTX 5050 in synthetic benchmarks in the Hermitage Akihabara review?

Hermitage Akihabara found the RX 9050 was about 18% slower than the RTX 5050 in synthetic benchmarks. The review also noted the worst losses for AMD were in ray tracing tests, and that some comparisons were skewed because the RTX 5050 was run with 4x or 6x MFG while the RX 9050 used standard FSR.

Q: How much slower was the RX 9050 than the RTX 5050 in the gaming benchmarks reported?

In the gaming benchmarks reported, the RX 9050 delivered about 30% less performance than the RTX 5050. This figure excludes the Doom benchmarks, which were noted as unreliable due to frame generation and memory issues.

Q: Why are some Doom benchmarks in the review considered unreliable for comparing the two cards?

They were unreliable because the NVIDIA card was allowed to use multi-frame generation (MFG) at 4x or 6x in some Doom tests while the AMD RX 9050 was tested with standard FSR, so the comparison was not level. At higher resolutions the RTX 5050's frame rate also crashed due to memory issues when frame generation was used, making those Doom results inconsistent and not a fair basis for comparison.

Q: What caveats did the article suggest before drawing firm conclusions from this single RX 9050 review?

The article warned that the review is skewed because the RTX 5050 was tested using 4x or 6x NVIDIA frame generation (MFG) in some cases while the RX 9050 was only run on standard FSR, so it was not a level playing field. It also noted Doom results were excluded where NVIDIA frame-generation caused memory-related frame drops, and that this is only a single source so more benchmarks and comparisons are needed before drawing firm conclusions.

Namely the fact that in some of the games tested, the NVIDIA graphics card got access to 4x or 6x MFG, whereas the AMD GPU was on standard FSR.

Darren Allan

Tech Reporter

Published Jul 31, 2026 8:40 AM CDT

2-minute read time

TL;DR: The first review of AMD's RX 9050 by a Japanese tech site shows that the GPU trails NVIDIA's RTX 5050 by around 18% in synthetic tests and 30% in gaming benchmarks. However, the testing is skewed as the Nvidia GPU ran 4x/6x MFG while the AMD model used standard FSR, at least in some cases. But still, the picture doesn't look that promising for the Radeon graphics card, especially given that it costs more.

Voice: DefaultSpeed

0:00 / 2:50

An initial review of AMD's new RX 9050 has popped up and the GPU doesn't fare that well in comparison to NVIDIA's RTX 5050, at least not at first glance - although there are issues with this evaluation as we'll see.

The RX 9050 has just arrived on the scene, and the 8GB graphics card has been tested by Japanese tech site Hermitage Akihabara (as flagged by VideoCardz).

Hermitage Akihabara put these AMD and NVIDIA GPUs through their paces in a head-to-head, comparing synthetic and real-world gaming benchmarks across a raft of titles.

There's a lot to unpick, but the upshot is that the RX 9050 was 18% off the pace of the RTX 5050 in synthetic tests, but it was a good deal slower with gaming benchmarks, with 30% less performance on offer from AMD. This excludes the Doom benchmarks that were clearly a bit all over the place, because at higher resolutions, the frame rate of the NVIDIA GPU cratered due to memory issues (with the frame generation being used).

And while this looks like a major stumble for AMD, in fact, you can't take some of these benchmarks at face value due to the use of the mentioned frame generation. The testers ran the RTX 5050 at 4x or 6x MFG in some cases, whereas that wasn't the case with the AMD GPU which used standard FSR - so this clearly isn't a level playing field.

Also, the worst losses for AMD in the synthetic benchmarks were ray tracing tests (as you might guess), but it's still not a great overall picture which is painted of the RX 9050. Especially considering that it retails for more than the RTX 5050 in the US (AMD has launched the new GPU at $279 compared to the $249 MSRP for the low-end NVIDIA rival).

Obviously we do need to wait for more benchmarks to be run and comparisons to be drawn, as we can't go off a single source anyway, especially given the way things have been skewed here with FG.

It's worth noting that where AMD does have a victory is with power usage, as the Radeon graphics card drew 10% less power idling (70W) and 16% less under load (182W), so that's a useful undercutting of the NVIDIA model.

Frequently Asked Questions

TweakBot answers common questions about this news using TweakTown's own coverage from this page and related content from our archive. Tap a question to reveal the answer, or type your own below.

Question #1

How did the RX 9050 compare to the RTX 5050 in synthetic benchmarks in the Hermitage Akihabara review?

Question #2

How much slower was the RX 9050 than the RTX 5050 in the gaming benchmarks reported?

Question #3

Why are some Doom benchmarks in the review considered unreliable for comparing the two cards?

Question #4

What caveats did the article suggest before drawing firm conclusions from this single RX 9050 review?

Have a question not listed here? Ask below and TweakBot will answer it.

As well as the 8GB incarnation of AMD's RX 9050, there is also a considerably watered down 4GB model, which is further hampered by its memory bus being cut in half. This variant is seemingly for OEM use only, and likely to be present in very cheap pre-built PCs.