dev: fixed an issues with CMP170

This commit is contained in:
empresa 2025-10-15 17:39:38 +07:00
parent 75df04553d
commit b8ef86c020
1 changed files with 11 additions and 2 deletions

View File

@ -241,6 +241,11 @@ def get_bus_spec(bus_id):
return PCIBusInfo()
def get_gpu_info():
GPU_ID_TO_NAME = {
"0x20C210DE": "NVIDIA CMP 170HX",
"0x208210DE": "NVIDIA CMP 170HX"
}
gpu_str = "0x Unknown"
nvml_err = False
gpu_mem = 0
@ -257,7 +262,7 @@ def get_gpu_info():
pass
nvidia_smi_return_code, nvidia_smi_stdout, nvidia_smi_stderr = utils.run_command(f"nvidia-smi --query-gpu=index,name,uuid,serial,memory.total --format=csv")
nvidia_smi_xl_return_code, nvidia_smi_xl_stdout, nvidia_smi_xl_stderr = utils.run_command("nvidia-smi --query-gpu=timestamp,name,pci.bus_id,driver_version,pstate,pcie.link.gen.max,pcie.link.gen.current,temperature.gpu,utilization.gpu,utilization.memory,memory.total,memory.free,memory.used --format=csv")
nvidia_smi_xl_return_code, nvidia_smi_xl_stdout, nvidia_smi_xl_stderr = utils.run_command("nvidia-smi --query-gpu=timestamp,name,pci.bus_id,driver_version,pstate,pcie.link.gen.max,pcie.link.gen.current,temperature.gpu,utilization.gpu,utilization.memory,memory.total,memory.free,memory.used,pci.device_id --format=csv")
if "Failed to initialize NVML" in nvidia_smi_stdout or "Failed to initialize NVML" in nvidia_smi_stderr or "Failed to initialize NVML" in nvidia_smi_xl_stdout or "Failed to initialize NVML" in nvidia_smi_xl_stderr:
nvml_err=True
@ -267,10 +272,14 @@ def get_gpu_info():
for index, line in enumerate(lines_xl):
parts = [s.strip() for s in line.split(',')]
if len(parts)>12 and index>0:
gpu_name = parts[1]
if gpu_name == "NVIDIA Graphics Device" and parts[13] in GPU_ID_TO_NAME:
gpu_name = GPU_ID_TO_NAME[parts[13]]
xl_gpu_info={
"id":index-1,
"timestamp": parts[0],
"name": parts[1],
"name": gpu_name,
"pcie_bus": parts[2].split(':', 1)[1],
"driver": parts[3],
"pstate": parts[4],