执行rocm-smi获取显卡设备信息,执行时发现在获取SCLK和功耗两个参数时报错,可能是SDK和显卡兼容方面的问题,不过没有影响后面的简单测试,所以也就没有理会。 AMD一直提倡一种混合异构架构(HSA),这种架构下CPU和GPU乃至各种异构算力被一视同仁,被纳入统一的编程模型开发,从rocminfo工具的输出可以看到这一点,每一个异构...
RSMI_DEV_PERF_LEVEL_DETERMINISM : rocm_smi.h RSMI_DEV_PERF_LEVEL_HIGH : rocm_smi.h RSMI_DEV_PERF_LEVEL_LOW : rocm_smi.h RSMI_DEV_PERF_LEVEL_MANUAL : rocm_smi.h RSMI_DEV_PERF_LEVEL_STABLE_MIN_MCLK : rocm_smi.h RSMI_DEV_PERF_LEVEL_STABLE_MIN_SCLK : rocm_smi.h RSMI_DE...
& rocm-smi" Mon Feb 24 14:44:01 2025 +---+
sclk_freq_limits : rsmi_od_volt_freq_data_t sdma_usage : rsmi_process_info_t status : rsmi_retired_page_record_t - t - temp_metric : id time_enabled : rsmi_counter_value_t time_running : rsmi_counter_value_t transfer_rate : rsmi_pcie_bandwidth_t type : rsmi_utilization...
rocm_smi_lib) for the installation guide.15 14 LD_LIBRARY_PATH should be set to the folder containing librocm_smi64. 16 15 17 16 ## Version @@ -220,7 +219,7 @@ $ sudo /opt/rocmbin/rocm-smi --setextremum max sclk 2100 220 219 DAMAGES CAUSED BY USE OF AMDGPU OUT...
I found only sclk is configurable. Is this expected, or did I miss anything? Thanks!rakataprime commented Jan 19, 2023 did u set the feature mask and performance to manual like ? rocm-smi --setperflevel manual sudo rocm-smi --setvc 2 1701 915 --autorespond y sudo rocm-smi --se...
也可使用类似nvidia-smi的rocm-smi命令查看显卡信息,就是输出内容有些简陋: ===ROCm System Management Interface=== === GPU Temp AvgPwr SCLK MCLK Fan Perf PwrCap VRAM% GPU% 0 54.0c 122.233W 1380Mhz 2000Mhz 38.82% auto 170.0W 9% 75% ===...
After rocm-smi -d 1 --setmclk 2 GPU Temp AvgPwr SCLK MCLK Fan Perf SCLK OD MCLK OD 1 38c 32.153W 300Mhz 2000Mhz 0.0% manual 0% 0% 0 36c 34.176W 300Mhz 2000Mhz 0.0% manual 0% 0% About HIP examples gpu-burn it keeps doing dimension 512x512 matrix mul, like A * B =C, an...
After rocm-smi -d 1 --setmclk 2 GPU Temp AvgPwr SCLK MCLK Fan Perf SCLK OD MCLK OD1 38c 32.153W 300Mhz 2000Mhz 0.0% manual 0% 0%0 36c 34.176W 300Mhz 2000Mhz 0.0% manual 0% 0% About HIP examples gpu-burn it keeps doing dimension 512x512 matrix mul, like A * B =C, and ...
I updated my Ubuntu 20.04 container to ROCm 3.9 from 3.8, and rocm-smi now produces the following error: ERROR:root:ROCm SMI returned 8 (the expected value is 0) Running rocm_smi.py produces the same error, but rocm_smi_deprecated.py see...