vault backup: 2023-09-27 11:41:52

4a1s/CP/PL - Aula 3.md

@@ -6,7 +6,16 @@ Logically following the multiplication process of matrices, C and A have spacial
 ### b)
 cache size = 25600 KB
 cache line/alignment= 64 B
+double -> 8
 
-n² misses 
+(64/8) = 8 elements
 
+**Estimation of misses:**
+- matrix C:   n² / 8
+- matrix A:   n² 
+
+
+>[!info]- Commands ran
+>nano /proc/cpuinfo
+>srun --partition=cpar perf cat /proc/cpuinfo
 ### c)