Benchmarks
Let’s move on to the benchmarks. For capacity reasons, I have actually opted for one game here, but it is definitely representative of the kind of performance you get. The game in question is Cyberpunk 2077, but it’s worth noting that the game couldn’t start when I was running the NVIDIA card. I got a ray tracing error when starting the game, which ended the game immediately. I first had to deactivate this card via the Device Manager so that I could start and play the game with the AMD card. I played once in QHD and in FHD to see what the AMD card could deliver and what my 3050 laptop GPU could manage. It was important to me to achieve a usable FPS number that was also suitable for gaming, which would be 60 FPS. I used the presets, but FSR 2 and DLSS were always used in a balanced way.
We use the graphics card tests from Igor’sLabs Workbench as a comparison. The normal Intel system from igor’sLAB was used with a Core i9-13900K overclocked to 6 GHz, 32 GB RAM (2x 16 GB) in the form of the Corsair Dominator Platinum DDR5 6400 (XMP profile active) on an MSI MEG Z690 Unify. Two 2TB NVMe SSDs from Corsair and MSI are used, as well as an AiO compact water cooling system for the CPU based on the Alphacool Eisbär Pro.4
Full HD with 1920 x 1080 pixels
What is clearly noticeable is that the eGPU does not come close to the performance of a desktop PC. The RX 7600 XT has 160.4 percent more FPS in the Workbench at Low with FSR. Things don’t look good for Medium either, with a difference of 151.4 %. It should be noted here that the components differ and that different CPUs also result in different conditions. However, you can see the increase compared to the RTX 3050. At low settings without FSR, there is an increase of 56.7 percent between eGPU and dGPU. Especially in the middle settings without DLSS or FSR, there is a difference of 77.8%. Here you can clearly see once again that the eGPU provides a good remedy in terms of performance. With the P1 1%, the change is pretty much the same for everyone.
In terms of latency, however, the difference is noticeable. While we have low latencies in the PC tests, there is a difference to be seen. Compared to the low settings without FSR or DLSS, there is a difference of 134.7% between eGPU and PC. I should note, however, that the latencies of the eGPU are quite moderate and definitely manageable. But it’s the same pattern here. The eGPU offers better latency values than the dGPU. The dGPU also has a higher latency of 57.9 percent in the low settings than the eGPU in comparison.
In FHD, I was easily able to get above 60 FPS. Here you can clearly see again that QHD offers some performance losses. But what has to be mentioned here is the fact that it represents a significant increase in performance compared to my RTX 3050 laptop GPU. I have up to 30 FPS more performance, which can make a big difference. And in the end, an RX7600 would certainly have been enough, because the transfer to the eGPU is severely limited. But I want to continue using the card in my own PC soon. Of course, these are fixed presets, because if you make a few more settings, you’ll certainly get a better FPS number, but I’ll leave it that way for now
21 Antworten
Kommentar
Lade neue Kommentare
Veteran
Urgestein
Mitglied
1
Veteran
1
Urgestein
Urgestein
Urgestein
1
Mitglied
Mitglied
Urgestein
Mitglied
Mitglied
Mitglied
Veteran
Urgestein
Veteran
Alle Kommentare lesen unter igor´sLAB Community →