Public bug reported: On a fresh, fully updated Ubuntu 26.10 development installation, CS2 ran at approximately 30–50 FPS with severe stuttering. Changing the desktop display refresh rate from 144 Hz to 60 Hz restored approximately 200–300+ FPS and normal GPU memory clocking. Returning to 144 Hz preserved the improved performance.
Environment: - Kernel: 7.3.0-8-generic - GPU: Sapphire Radeon RX 6700 (Navi 22), PCI ID 1002:73df, subsystem 1da2:e445 - CPU: Ryzen 7 5700X3D - Mesa: 26.2.4-1ubuntu1 - linux-firmware: 20260915.git1522c78a.1ubuntu - GNOME Wayland - Display: GPR24C1MS144, DisplayPort DP-1, 1920x1080 at approximately 144 Hz, scale 1 - Steam installed using its Debian package - CS2: 1920x1080 borderless, VSync off, 4x MSAA, fps_max 400 During the slowdown, repeated live readings showed: - GPU utilization: 99% - pp_dpm_mclk: lowest state selected, 96 MHz - hwmon freq2_input: 96000000 Hz - GPU core clock: approximately 2.65 GHz - GPU power: approximately 65–71 W - GPU temperature: approximately 44–50 C; hotspot 52–59 C Diagnostic attempts: 1. Temporarily selected power_dpm_force_performance_level=high. The driver reported high, but repeated readings still showed 96 MHz memory clock. Performance did not improve. 2. Temporarily selected manual control and wrote 3 to pp_dpm_mclk, the highest listed memory state. The printed table still showed 96 MHz selected, and performance remained poor. No command error was reported. Both temporary tests restored automatic control afterward. Working recovery: 1. While CS2 remained running, changed desktop refresh rate from 144 Hz to 60 Hz in Ubuntu Settings. 2. Performance immediately improved to approximately 200–300+ FPS. 3. Live readings confirmed memory state 3 selected at 1000 MHz and freq2_input=1000000000 Hz. GPU power rose to approximately 136–139 W. 4. Changed back to 144 Hz. Performance remained good, memory stayed at 1000 MHz, and automatic GPU control remained active. Expected: GPU memory should clock up appropriately under gaming load without requiring a display-mode toggle. Reproducibility: Observed once. The original trigger is unknown. Recurrence after reboot or suspend has not yet been tested, and no older-kernel comparison has been performed. A kernel regression is therefore not established. The original sampled display mode used fixed refresh. Variable refresh rate was enabled only after recovery. An older, potentially related report describes memory clock stuck at the lowest frequency with kernel 6.4, marked fixed in 6.5: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1042503 It was forwarded to: https://gitlab.freedesktop.org/drm/amd/-/issues/2657 I cannot confirm that this is the same underlying defect. ProblemType: Bug DistroRelease: Ubuntu 26.10 Package: linux-image-7.3.0-8-generic 7.3.0-8.8 ProcVersionSignature: Ubuntu 7.3.0-8.8-generic 7.3.0-rc5 Uname: Linux 7.3.0-8-generic x86_64 ApportVersion: 2.36.0-0ubuntu1 Architecture: amd64 AudioDevicesInUse: USER PID ACCESS COMMAND /dev/snd/controlC1: ilker 4310 F.... wireplumber /dev/snd/controlC0: ilker 4310 F.... wireplumber /dev/snd/seq: ilker 4306 F.... pipewire CasperMD5CheckResult: pass CurrentDesktop: ubuntu:GNOME Date: Tue Oct 6 22:33:33 2026 InstallationDate: Installed on 2026-10-06 (0 days ago) InstallationMedia: Ubuntu 26.10 "Stonking Stingray" - Daily amd64 (20261005) MachineType: Gigabyte Technology Co., Ltd. B550 GAMING X V2 ProcEnviron: LANG=tr_TR.UTF-8 PATH=(custom, no user) SHELL=/bin/bash TERM=xterm-256color XDG_RUNTIME_DIR=<set> ProcFB: 0 amdgpudrmfb ProcKernelCmdLine: BOOT_IMAGE=/boot/vmlinuz-7.3.0-8-generic root=UUID=b7dd3cd1-7487-4610-a00f-d5d693b306d7 ro quiet splash crashkernel=2G-4G:320M,4G-32G:512M,32G-64G:1024M,64G-128G:2048M,128G-:4096M RfKill: SourcePackage: linux UpgradeStatus: No upgrade log present (probably fresh install) dmi.bios.date: 04/14/2026 dmi.bios.release: 5.17 dmi.bios.vendor: American Megatrends International, LLC. dmi.bios.version: F21a dmi.board.asset.tag: Default string dmi.board.name: B550 GAMING X V2 dmi.board.vendor: Gigabyte Technology Co., Ltd. dmi.board.version: Default string dmi.chassis.asset.tag: Default string dmi.chassis.type: 3 dmi.chassis.vendor: Default string dmi.chassis.version: Default string dmi.modalias: dmi:bvnAmericanMegatrendsInternational,LLC.:bvrF21a:bd04/14/2026:br5.17:svnGigabyteTechnologyCo.,Ltd.:pnB550GAMINGXV2:pvrDefaultstring:rvnGigabyteTechnologyCo.,Ltd.:rnB550GAMINGXV2:rvrDefaultstring:cvnDefaultstring:ct3:cvrDefaultstring:skuDefaultstring:pfaB550MB: dmi.product.family: B550 MB dmi.product.name: B550 GAMING X V2 dmi.product.sku: Default string dmi.product.version: Default string dmi.sys.vendor: Gigabyte Technology Co., Ltd. ** Affects: linux (Ubuntu) Importance: Undecided Status: New ** Tags: amd64 apport-bug stonking wayland-session -- You received this bug notification because you are a member of Ubuntu Bugs, which is subscribed to Ubuntu. https://bugs.launchpad.net/bugs/2169747 Title: amdgpu: RX 6700 memory clock stuck at 96 MHz under load until display refresh rate is toggled To manage notifications about this bug go to: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/2169747/+subscriptions -- ubuntu-bugs mailing list [email protected] https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
