# NVIDIA Grace (ARM Neoverse V2) Cache ## Cache hierarchy | Level | Present | SLeeLa model | |---|---|---| | L1 instruction | Yes | present | | L1 data | Yes | present | | L2 | Yes | present | | L3 | Yes | present | ## Timing consequence The model records the presence of the cache levels above. The functional model accesses memory through the bus interface; cache levels are represented as architectural facts and timing hints rather than a cycle-accurate replacement policy. ## Variant rule Do not attach cache objects that this part does not actually contain, and do not remove levels it does. A different SKU or derivative with a different cache configuration is a separate profile. ## Sources - https://www.nvidia.com/en-us/data-center/grace-cpu/ - https://en.wikipedia.org/wiki/ARM_Neoverse