Infrastructure · Infrastructure resource

GPU Compute Instance

Infrastructure resourceInfrastructureInfrastructurearc:GPUComputeInstance

A compute-isolated subdivision of a GPU partition that owns a subset of the partition's streaming multiprocessors while sharing its memory pool with sibling compute instances.

Responsibility. Gives memory-heavy, compute-light workloads compute separation inside one partition's shared memory.

Also known as: Compute Instance (CI), MIG compute instance

Direct neighbourhood (hover for relationship types)

Relationships

hosts structural

Self-Hosted GPU Embedding Service Ch7.6

is configured by structural

GPU Partition Layout abstract Ch7.6

Design guidance

MAY host several embedding services inside one GPU partition, since they need memory for weights but little compute.

Classification

Technologies: NVIDIA Multi-Instance GPU (MIG)
Quality attributes: Performance efficiency (ISO/IEC 25010)

Sources

Ch7.6: T. Nguyen, "Multi-Instance GPU," in Mastering Agentic AI Systems: Guide for the NVIDIA NCP-AAI Exam, 1st ed. 2026, ch. 7.6. ISBN: 9798244538229.