12 Hours, 7 Layers, 1 Hard Wall: Debugging Memory on DGX Spark

This site does not allow itself to be embedded. You can still read it on the original site — the toolbar below keeps your place in the directory.

Yesterday I spent 12 consecutive hours debugging why MAX Engine allocates 117GB of RAM on my DGX Spark, regardless of what limits I set. I traced the problem through seven system layers, wrote six patches, and ultimately hit a hard wall in the kernel. The conclusion : On unified memory architectures, there is no userspace solution. The NVIDIA kernel module allocates directly from physical memory,…

Read on lab-stack.com ↗