Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

nvidia mode is choppy while hybrid mode appears to have more performance. #498

Open
DrunkenToast opened this issue Jun 5, 2022 · 1 comment

Comments

@DrunkenToast
Copy link

Describe the bug
Performance of the nvidia gpu is slower when ran with prime-run than my amd integrated gpu in hybrid mode.
However in dedicated nvidia mode it performs better.

I know some overhead is to be expected but this difference seems to be huge. Any mistake on my end?

System info
EndeavourOS
KDE 5.24.5
SDDM
Optimus Manager v1.4

Nvidia GPU: NVIDIA GeForce RTX 2060/PCIe/SSE2 (Mobile)
Driver Version: 515.43.04 CUDA Version: 11.7

CPU/iGPU: Amd Ryzen 7 4800H | AMD RENOIR (LLVM 13.0.1, DRM 3.46, 5.18.1-arch1-1)

Custom optimus-manager:

[amd]
DRI=3
driver=amdgpu
tearfree=

[intel]
DRI=3
accel=
driver=modesetting
modeset=yes
tearfree=

[nvidia]
DPI=96
PAT=yes
allow_external_gpus=no
dynamic_power_management=fine
dynamic_power_management_memory_threshold=0
ignore_abi=no
modeset=yes
options=overclocking

[optimus]
auto_logout=yes
pci_power_control=no
pci_remove=no
pci_reset=no
startup_auto_battery_mode=integrated
startup_auto_extpower_mode=nvidia
startup_mode=hybrid
switching=none

Logs

Optimus Manager (Client) version 1.4

Current GPU mode : hybrid
GPU mode requested for next login : no change
GPU at startup : hybrid
Temporary config path: no

Benchmarks

Hybrid iGPU amd:

=======================================================
    glmark2 2021.12
=======================================================
    OpenGL Information
    GL_VENDOR:      AMD
    GL_RENDERER:    AMD RENOIR (LLVM 13.0.1, DRM 3.46, 5.18.1-arch1-1)
    GL_VERSION:     4.6 (Compatibility Profile) Mesa 22.1.0
    Surface Config: buf=32 r=8 g=8 b=8 a=8 depth=24 stencil=0
    Surface Size:   800x600 windowed
=======================================================
[build] use-vbo=false: FPS: 9009 FrameTime: 0.111 ms
[build] use-vbo=true: FPS: 11434 FrameTime: 0.087 ms
[texture] texture-filter=nearest: FPS: 9172 FrameTime: 0.109 ms
[texture] texture-filter=linear: FPS: 9194 FrameTime: 0.109 ms
[texture] texture-filter=mipmap: FPS: 8689 FrameTime: 0.115 ms
[shading] shading=gouraud: FPS: 7909 FrameTime: 0.126 ms
[shading] shading=blinn-phong-inf: FPS: 7385 FrameTime: 0.135 ms
[shading] shading=phong: FPS: 7280 FrameTime: 0.137 ms
[shading] shading=cel: FPS: 6927 FrameTime: 0.144 ms
[bump] bump-render=high-poly: FPS: 4851 FrameTime: 0.206 ms

Hybrid nvidia offloading with prime-run:
prime-run glmark2

=======================================================
    glmark2 2021.12
=======================================================
    OpenGL Information
    GL_VENDOR:      NVIDIA Corporation
    GL_RENDERER:    NVIDIA GeForce RTX 2060/PCIe/SSE2
    GL_VERSION:     4.6.0 NVIDIA 515.43.04
    Surface Config: buf=32 r=8 g=8 b=8 a=8 depth=24 stencil=0
    Surface Size:   800x600 windowed
=======================================================
[build] use-vbo=false: FPS: 1968 FrameTime: 0.508 ms
[build] use-vbo=true: FPS: 2450 FrameTime: 0.408 ms
[texture] texture-filter=nearest: FPS: 2443 FrameTime: 0.409 ms
[texture] texture-filter=linear: FPS: 2428 FrameTime: 0.412 ms
[texture] texture-filter=mipmap: FPS: 2412 FrameTime: 0.415 ms
[shading] shading=gouraud: FPS: 2410 FrameTime: 0.415 ms
[shading] shading=blinn-phong-inf: FPS: 2417 FrameTime: 0.414 ms
[shading] shading=phong: FPS: 2384 FrameTime: 0.419 ms
[shading] shading=cel: FPS: 2388 FrameTime: 0.419 ms
[bump] bump-render=high-poly: FPS: 2278 FrameTime: 0.439 ms

Nvidia mode:

=======================================================
    glmark2 2021.12
=======================================================
    OpenGL Information
    GL_VENDOR:      NVIDIA Corporation
    GL_RENDERER:    NVIDIA GeForce RTX 2060/PCIe/SSE2
    GL_VERSION:     4.6.0 NVIDIA 515.43.04
    Surface Config: buf=32 r=8 g=8 b=8 a=8 depth=24 stencil=0
    Surface Size:   800x600 windowed
=======================================================
[build] use-vbo=false: FPS: 4916 FrameTime: 0.203 ms
[build] use-vbo=true: FPS: 12187 FrameTime: 0.082 ms
[texture] texture-filter=nearest: FPS: 11470 FrameTime: 0.087 ms
[texture] texture-filter=linear: FPS: 11570 FrameTime: 0.086 ms
[texture] texture-filter=mipmap: FPS: 11874 FrameTime: 0.084 ms
[shading] shading=gouraud: FPS: 9848 FrameTime: 0.102 ms
[shading] shading=blinn-phong-inf: FPS: 10904 FrameTime: 0.092 ms
[shading] shading=phong: FPS: 10264 FrameTime: 0.097 ms
[shading] shading=cel: FPS: 10965 FrameTime: 0.091 ms
[bump] bump-render=high-poly: FPS: 9351 FrameTime: 0.107 ms
@heddxh
Copy link

heddxh commented Nov 10, 2022

Without Optimus-manager still encounters this problem, for me it is i+n graphic cards. Tested by glmark2, Intel gets 3500 compared to 1500 on prime-run glmark2 .
So maybe it isn’t Optimus-manager’s problem.

@nwildner nwildner changed the title NVIDIA offloading slower than integrated (only hybrid) nvidia mode is choppy while hybrid mode appears to have more performance. Jul 2, 2024
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

4 participants