Identifying bottleneck of Razer Blade 2019 + Core v1 + RTX 2080Ti
I have been trying to identify issue with performance of my ASUS TURBO GeForce RTX 2080Ti 11GB enclosed in Razer Core V1 and connected to Razer Blade 2019 RTX 2080.
I know there is perf hit while using TB3 eGPU setup but right now RTX 2080MQ inside Blade itself over-performs the external RTX 2080Ti by significant amount of FPS.
For example in R6 Siege benchmark I can see nearly 100 FPS difference in favor of internal 2080. In BFV the difference is small but still its about ~20FPS (depends on map, situation etc) using ultra settings with DXR. The benchmarks are not precise, internal 2080 was outputting to 1080p internal 144Hz display and 2080Ti to 1440p external 165Hz display, but still the different shouldn't be that big, right?
My first thoughts went to PSU inside Core v1 which should provide 350W and based on this specs https://www.techpowerup.com/gpu-specs/asus-turbo-rtx-2080-ti.b6281 the TDP of card is 250W, which should be enough.
Is this expected behavior or maybe the PSU is not sufficient? Maybe Razer Core X with better PSU which should provide 500W to card, could solve this issue?
Thanks for advice.
Firstly, the Razer Core has a max GPU power delivery of 375W and your 2080Ti will pull at most 260W without overclocking, so your issue is not there. Next, I think you have underestimated the difference in performance between playing on a 1080p display and a 1440p display. According to this thread running 1440p over 1080p requires about 1.8x more GPU power. This ontop of the ~20% performance drop that is standard for TB3 eGPUs would likely account for the FPS drop you are experiencing.
You could probbably test the 2080MQ on the external display to see how close in perfomance the two cards are running the same resolution.
You are right, I did not expect the perf hit between 1080p and 1440p to be that high.
Also, I have noticed the 2080MQ is using DLSS, which 2080Ti is not and it cannot be enabled for 1440p, which for sure helps to increase framerate too.
Generally in BFV with 2080Ti I see very similar results mentioned here https://www.hardocp.com/article/2019/01/07/battlefield_v_nvidia_ray_tracing_rtx_2080_ti_performance/4
And my theory regarding the Siege - the frame rate is mostly above 120 FPS and for sure 2080Ti would handle much higher frame rates but I suppose here I hit TB3 tlimitations.
To summarize this - everything seems to be working correctly as expected 🙂
Thank you for your help.