Hi everyone:
The nano has 4G LPDDR4 memory. I wonder how much memory the gpu has?Any command tools?
When running the algorithm, the display is about 500M idle, which is far from enough.How do I allocate more memory to the gpu?
Best regards!
Hi everyone:
The nano has 4G LPDDR4 memory. I wonder how much memory the gpu has?Any command tools?
When running the algorithm, the display is about 500M idle, which is far from enough.How do I allocate more memory to the gpu?
Best regards!
Hi,
You should be able to allocate most system RAM except kernel-reserved memory:
[url]https://devtalk.nvidia.com/default/topic/1002349/jetson-tx2/jetson-tx2-gpu-memory-/post/5119935/#5119935[/url]
You could stop X server with
sudo service gdm stop
If you would run a code from https://devtalk.nvidia.com/default/topic/491518/cuda-programming-and-performance/cudamemgetinfo-how-does-it-work-33-/post/3522842/#3522842, you’ll see that in a system with stopped gdm
, ~800MB of GPURAM are still consumed by unknown things.