Hey folks, Been daily-driving OrangePi 6 / 6B with latest Armbian noble-edge (6.12-rc kernel + rknpu 0.9.9-8) for local LLM inference. Even after manually loading the closed-source RKNPU DDK 0.9.9, switching to the "new" rknpu.ko from Rockchip’s 5.10 BSP, and forcing INT4 quantization via llama.cpp built with -DRKNN_RT=ON, the NPU utilization reported by rk_nn_tool tops out around 36-42% when running Q4_K_M 70B-class models (tokens/s barely hits ~19-21 t/s). CPU is almost idle, no thermal throttling, 16GB LPDDR4X fully available. Is this still the known "18 TOPS theoretical vs real ~7-8 TOPS usable" ceiling, or did anyone manage to push past 65%+ utilization on mainline-ish kernels in late 2025? Happy to share my build scripts and rk_nn_tool logs if anyone wants to dig deeper. Bonus question: has anyone successfully coerced the NPU into handling 2+ concurrent contexts without dropping to zero utilization (the infamous "second model load kills the first" bug)?

ok same error. The maxio module works with this snippet though : mdio { compatible = "snps,dwmac-mdio"; #address-cells = <0x01>; #size-cells = <0x00>; phandle = <0xee>; ethernet-phy@0 { compatible = "ethernet-phy-id7b74.4411", "ethernet-phy-ieee802.3-c22" reg = <0x00>; status = "okay"; phandle = <0x51>; }; };

yes. Things that can be tried: - DTS phy-mode = "rgmii-id"; or phy-mode = "rgmii-rxid"; - UBOOT CONFIG_PHY_REALTEK=y CONFIG_DWC_ETH_QOS=y CONFIG_DWC_ETH_QOS_ROCKCHIP=y CONFIG_DM_PMIC=y CONFIG_PMIC_RK8XX=y CONFIG_REGULATOR_RK8XX=y CONFIG_PWM_ROCKCHIP=y

OK do you mean boot with ethernet working using the maxio module then change DTB and reboot?

Can you boot in a working ethernet and then reboot to this trial?