cache-config
cache-config
Set the cache hierarchy to collect
modeling data for CPU cache behavior during Trip Counts & FLOP analysis.
GUI Equivalent
Syntax
--cache-config
=
<string>
Arguments
<string>
follows this template:
[num_of_level1_caches]:[num_of_ways_level1_connected]:[level1_cache_size]:[level1_cacheline_size]/
[num_of_level2_caches]:[num_of_ways_level2_connected]:[level2_cache_size]:[level2_cacheline_size]/
[num_of_level3_caches]:[num_of_ways_level3_connected]:[level3_cache_size]:[level3_cacheline_size]
For example: 4:8w:32k:64l/4:4w:256k:64l/1:16w:6m:64l
Usage
When no specific configuration is set, the
Intel Advisor
uses system cache hierarchy for modeling.
Cache simulation modeling applies to the following:
- Memory Access Patterns analysis - This basic simulation functionality models accurate memory footprints, miss information, and cache line utilization for a downstream Memory Access Patterns report.
- CPU / Memory Roofline Insightsperspective - This enhanced simulation functionality models multiple levels of cache for a downstream Memory-Level Roofline chart or Roofline interactive HTML report.
This option is applicable only to Trip Counts and FLOP
and Roofline analyses.
Example
- Run a Survey analysis.
- Run a Trip Counts & FLOP analysis. Model cache behavior for the specified configuration.
advisor --collect=survey
--project-dir=./advi_results -- ./myApplication
advisor --collect=tripcounts --flop
--enable-cache-simulation
--cache-config=4:8w:32k:64l/4:4w:256k:64l/1:16w:6m:64l
--project-dir=./advi_results -- ./myApplication
Run Roofline analysis for all memory levels
(Memory-Level Roofline) for the specified cache configuration.
advisor
--collect=roofline --enable-cache-simulation
--cache-config=4:8w:32k:64l/4:4w:256k:64l/1:16w:6m:64l
--project-dir=./advi_results -- ./myApplication