# GPU host pins, percentes-gpu 10.0.0.122 / [public address redacted], captured 2026-09-16T19:19:42Z
== nvidia-smi --query ==
name, driver_version, persistence_mode, clocks.current.sm [MHz], clocks.current.memory [MHz], clocks.max.sm [MHz], power.limit [W], power.max_limit [W], pstate
NVIDIA L40, 570.195.03, Enabled, 210 MHz, 405 MHz, 2490 MHz, 300.00 W, 300.00 W, P8
== cuda version.json ==
{
   "cuda" : {
      "name" : "CUDA SDK",
      "version" : "12.8.1"
   },
   "cuda_cccl" : {
      "name" : "CUDA C++ Core Compute Libraries",
      "version" : "12.8.90"
   },
   "cuda_cudart" : {
      "name" : "CUDA Runtime (cudart)",
      "version" : "12.8.90"
   },
   "cuda_cuobjdump" : {
      "name" : "cuobjdump",
      "version" : "12.8.90"
   },
   "cuda_cupti" : {
      "name" : "CUPTI",
      "version" : "12.8.90"
== docker ==
Docker version 29.1.3, build f52814d
== lsblk ==
NAME     SIZE ROTA MODEL
vda      100G    1 
├─vda1    99G    1 
├─vda14    4M    1 
├─vda15  106M    1 
└─vda16  913M    1 
vdb      725G    1 
== persistence mode attempt ==
Persistence mode is already Enabled for GPU 00000000:00:07.0.
All done.