arXiv:2606.11916v1 Announce Type: cross Abstract: This paper proposes an empirical methodology to study software aging in GPU-based LLM serving systems. Traditional aging studies focus on CPU-centric software with relatively regular workloads; LLM serving is different, spanning a Python host and a CUDA device, handling requests whose cost varies...
Read the full article at the source.
![[Ekstra] Sopra Steria: Hun er ny leder](https://images.gfx.no/1000/2928/2928034/Kristine%20Bolstad%20og%20Kjell%20Rusti.jpg)
Comments (0)
No comments yet. Be the first to comment!