Radeon Pro WX 4100 Test (Ryzen 5 Pro 2400G, Bazzite, Lenovo Thinkcentre M725s) Gaming Benchmarks
No Commentary
These aren't in-depth benchmarks, and I don't own a lot of the popular games people play. Just showing this as a test or my new Small Form Factor Linux PC. A fun side-project for myself.
I bought the Radeon WX 4100 for my HP Slimline, but the card cooked in that case, reaching over 90 degrees, Celsius even after fresh thermal paste, because there was no airflow in my HP slimline 270's cas. Instead, I found a Lenovo Thinkcentre 725s with a Ryzen 5 Pro 2400G, 16 gigs of DDR4 ram, and a 265 gig NVME for $110 on Ebay. I managed to pick up a 512 gig Sata SSD for around $64, and the Radeon Pro WX 4100 also cost me $64. Both came out to $64.64. I also bought two mini display-port to HDMI Adapters for around $6, which are required ad the WX 4100 only has 4-mini display port outs. The whole thing cost me a little over $245, which isn't bad. Capabilities-wise, I would put it below a base PS4. It depends on the game.
For Bazzite, I installed it with Game mode, but after struggling with configuring things the way I wanted, and realizing that the games for whatever reason, just don't run as well as they do in desktop mode, I put in the "Bruh" command in the terminal to Rebase my OS to change it to the desktop version of Bazzite instead. You can do this without having to reinstall Bazzite. That's pretty cool.
As for the Radeon Pro WX 4100, from the benchmark sites, it is, essentially, identical in performance to the RX 560 4GB. TechPowerup give the RX 560 a 6% lead in their standings, but the various benchmark sites I checked put the WX 4100 at less than 1% better than the RX 560. I don't have an RX 560 to test this though, so take everything with a grain of Salt. The RX 560 does have 1ghz higher memory clock and higher memory bandwith, so the benchmark sites could be wrong, or it could be that, in practice, the performance between the two cards is almost identical. I don't own an RX 560 to check.
The WX 4100 is a decent card. It's not mind blowing in performance, but it is better than the Vega 11 graphics on the Ryzen 5 Pro 2400G. APU's are still cool though, and were a godsend during the GPU shortage of 2020.
For Batman: Arkham Knight, I ran high settings. I'm pretty sure I will just cap the framerate at 30 fps to get the console experience, or I could drop the resolution to 900p or 720p and tweak settings to get it close to the 60 fps mark. I'm thinking a 30 fps cap would be fine as I play it with a controller anyway.
For Marvel's Spider-Man Remastered, I like native 720 over 1080p with upscaling. Upscaling adds some overhead which can cause a hit to the performance, and since I only have a 24" 1080p monitor, playing the game at 720p, and letting my monitor scale it to 1080p just looks better to me than FSR or Temporal Injection. For this one, I will use Mangohud, or Mango Juice to cap the framerate at 30fps for a smoother experience. It's averaging around 40fps, but it would smoother at a capped 30.
For Spyro, I wanted to see how Unreal Engine 4 ran on the card, and 1080p low get some good results. It does seem like, if I turn anything in the settings menu up even on notch higher, it drops to under 60 fps average at 1080p. I will probably play this at 900p medium with low shadows, or 720p mixture of medium to high settings for better visuals at a lower res with a decent framerate. I would still cap it at 60.
0:00 System Specs
0:06 DOOM 2016
2:00 Batman Arkham Knight
4:22 Spyro Reignited Trilogy
6:57 Doom Eternal
11:05 Marvel's Spider-Man Remastered
#bazzite #linuxgaming #radeon #radeonpro #pcbenchmarks