Arteris Articles

SemiWiki: A Last-Level Cache for SoCs

JP Loison, Senior Corporate Application Architect at Arteris IP, chats with Bernard Murphy in this SemiWiki blog:

A Last-Level Cache for SoCs  

July 19th,  2018 - By Bernard Murphy

2018-07-24-soc-diagram-codacache-600pxBased on a discussion with JP Loison, Bernard Murphy (SemiWiki) describes Arteris IP’s latest IP introduction - a last-level cache (LLC) for use outside the coherent domain, for accelerators and other IP that can benefit from cache support. He also touches on benefits offered by significant flexibility in adapting the IP for different use-models.

We tend to think of cache primarily as an adjunct to processors to improve performance. Reading and writing main memory (DRAM) is very slow thanks to all the package and board impedance between chips. If you can fetch blocks of contiguous memory from the DRAM to a local on-chip memory, locality of reference in most code ensures much faster access for many subsequent operations which will frequently find the data/addresses they need in these cached copies. This greatly improves overall performance, despite need at times to update cache contents from a different location (and maybe store back in DRAM what must be evicted from the cache).

To read the entire article, please click here:

Topics: SoC cache coherent IP CPU Ncore semiwiki cache CodaCache on-chip memory performance multi-processor systems congestion configurability last level cache scratchpad way partitioning