First Radeon RX 9050 benchmarks show 27% lower ray-tracing performance than RTX 5050
ASRock Challenger 8GB becomes the first card tested.
The first detailed Radeon RX 9050 review has been published by Japan’s Hermitage Akihabara. The outlet tested the ASRock Radeon RX 9050 Challenger 8GB against the GIGABYTE GeForce RTX 5050 WINDFORCE OC 8GB using a Ryzen 7 9800X3D system. Across the eight overall 3DMark scores, our calculation puts the RX 9050 around 18.2% behind the RTX 5050.
AMD launched the Radeon RX 9050 quietly for selected regions, with only a small number of partner models announced so far. ASRock listed Challenger cards with both 8GB and 4GB memory configurations. The model supplied to GDM for testing was the retail 8GB version, model RX9050 CL 8G.
© GDM Navi 44 XE
If you look closely at the BIOS data, you’ll notice that AMD has given this Navi 44 GPU a new name, the Navi 44 XE. The RX 9050 uses the Navi 44 GPU with 16 Compute Units, 1,024 Stream Processors, 16 Ray Accelerators and 32 AI Accelerators.
It has 8GB of GDDR6 memory running at 18 Gbps across a 128-bit interface, providing 288 GB/s of bandwidth. AMD specifies a 1,920 MHz game clock, 2,600 MHz boost clock and 92W board power.
Source: GDM
RTX 5050 leads most synthetic and gaming tests
The largest 3DMark deficits appeared in tests using ray tracing. The RX 9050 was 25.7% slower in Speed Way and 26.7% slower in Port Royal. The gap fell to 13.0% in Steel Nomad, 10.7% in Time Spy and 10.1% in Time Spy Extreme.
The RTX 5050 also led most game tests. The RX 9050 was 18.6% slower in Final Fantasy XIV at 1080p, 23.5% slower in Apex Legends and 14.6% slower in Rainbow Six Siege X using the Medium preset. The cards matched at 64 FPS in Rainbow Six Siege X at 4K with the Ultra+ preset.
Some of the larger differences do not represent direct GPU rendering comparisons. GDM enabled standard FSR frame generation on the Radeon card while the RTX 5050 used DLSS Multi Frame Generation at 4x or 6x in Cyberpunk 2077, Onimusha, DOOM and Assassin’s Creed Black Flag. This gave NVIDIA a large advantage at 1080p, although the RTX 5050 dropped to around 10 to 17 FPS in DOOM at 1440p and 4K due to memory limitations.
On average, RX 9050 is 18% slower than RTX 5050 in synthetic tests and 30% slower in gaming (if we exclude Doom, which obviously doesn’t work well at 4K with these GPUs).
What’s better from Radeon, it used less power in GDM’s testing. Total system consumption reached 182.4W under load with the RX 9050, compared with 217.5W for the RTX 5050. This is a reduction of 35.1W, or 16.1%. The test also recorded a peak GPU temperature of 56°C and a hotspot temperature of 69°C on the ASRock card.
| Radeon RX 9050 8GB vs GeForce RTX 5050 8GB | |||
|---|---|---|---|
| GDM Test | Radeon RX 9050 | GeForce RTX 5050 | Difference (%) |
| 3DMark | |||
| 3DMark Port Royal | 4531 | 6181 | -26.7% |
| 3DMark Speed Way | 1873 | 2520 | -25.7% |
| 3DMark Steel Nomad | 1985 | 2282 | -13.0% |
| 3DMark Fire Strike overall | 20700 | 26749 | -22.6% |
| 3DMark Fire Strike graphics | 21571 | 28062 | -23.1% |
| 3DMark Fire Strike Extreme overall | 10229 | 12920 | -20.8% |
| 3DMark Fire Strike Extreme graphics | 10287 | 13150 | -21.8% |
| 3DMark Fire Strike Ultra overall | 5286 | 6313 | -16.3% |
| 3DMark Fire Strike Ultra graphics | 5072 | 6042 | -16.1% |
| 3DMark Time Spy overall | 9441 | 10571 | -10.7% |
| 3DMark Time Spy graphics | 8866 | 10036 | -11.7% |
| 3DMark Time Spy Extreme overall | 4450 | 4952 | -10.1% |
| 3DMark Time Spy Extreme graphics | 4171 | 4701 | -11.3% |
| Gaming | |||
| Final Fantasy XIV, 1080p | 12895 | 15847 | -18.6% |
| Final Fantasy XIV, 1440p | 8136 | 10059 | -19.1% |
| Final Fantasy XIV, 4K | 3994 | 5102 | -21.7% |
| Apex Legends, 1080p | 147.1 | 192.4 | -23.5% |
| Apex Legends, 1440p | 101.4 | 128.2 | -20.9% |
| Apex Legends, 4K | 72.6 | 74.8 | -2.9% |
| Rainbow Six Siege X Medium, 1080p | 204 | 239 | -14.6% |
| Rainbow Six Siege X Medium, 1440p | 125 | 150 | -16.7% |
| Rainbow Six Siege X Medium, 4K | 65 | 73 | -11.0% |
| Rainbow Six Siege X Ultra+, 1080p | 129 | 141 | -8.5% |
| Rainbow Six Siege X Ultra+, 1440p | 116 | 126 | -7.9% |
| Rainbow Six Siege X Ultra+, 4K | 64 | 64 | 0.0% |
| Cyberpunk 2077 Medium, 1080p | 210.79 | 300.16 | -29.8% |
| Cyberpunk 2077 Medium, 1440p | 134.45 | 185.25 | -27.4% |
| Cyberpunk 2077 Medium, 4K | 64.27 | 107.47 | -40.2% |
| Cyberpunk 2077 Ultra, 1080p | 142.01 | 241.72 | -41.3% |
| Cyberpunk 2077 Ultra, 1440p | 87.51 | 147.39 | -40.6% |
| Cyberpunk 2077 Ultra, 4K | 40.42 | 77.19 | -47.6% |
| Onimusha Way of the Sword Medium, 1080p | 134.97 | 166.60 | -19.0% |
| Onimusha Way of the Sword Medium, 1440p | 103.40 | 143.15 | -27.8% |
| Onimusha Way of the Sword Medium, 4K | 62.09 | 103.55 | -40.0% |
| Onimusha Way of the Sword Ultra, 1080p | 110.07 | 149.66 | -26.5% |
| Onimusha Way of the Sword Ultra, 1440p | 86.43 | 128.51 | -32.7% |
| Onimusha Way of the Sword Ultra, 4K | 53.33 | 92.13 | -42.1% |
| DOOM The Dark Ages Medium, 1080p | 132.74 | 296.63 | -55.3% |
| DOOM The Dark Ages Medium, 1440p | 85.11 | 17.17 | +395.7% |
| DOOM The Dark Ages Medium, 4K | 39.71 | 9.87 | +302.3% |
| DOOM The Dark Ages Ultra Nightmare, 1080p | 114.17 | 280.13 | -59.2% |
| DOOM The Dark Ages Ultra Nightmare, 1440p | 73.79 | 10.71 | +589.0% |
| DOOM The Dark Ages Ultra Nightmare, 4K | 35.56 | 9.62 | +269.6% |
| Assassin’s Creed Black Flag Medium, 1080p | 94 | 164 | -42.7% |
| Assassin’s Creed Black Flag Medium, 1440p | 78 | 153 | -49.0% |
| Assassin’s Creed Black Flag Medium, 4K | 53 | 127 | -58.3% |
| Assassin’s Creed Black Flag Highest, 1080p | 57 | 135 | -57.8% |
| Assassin’s Creed Black Flag Highest, 1440p | 50 | 122 | -59.0% |
| Assassin’s Creed Black Flag Highest, 4K | 36 | 92 | -60.9% |
| System power | |||
| System power, idle | 70.5W | 78.9W | -10.6% |
| System power, load | 182.4W | 217.5W | -16.1% |
Source: GDM
