Added results for "target-cpu=native"

Author: jorgecarleitao

README.md

@@ -17,7 +17,47 @@ Algorithms implemented:
 * `nonsimd`: vertical sum over lanes with a reduce at the end using Rust arrays
 * `naive`: sum using rust iterators
 
-## Bench results on my computer
+## Bench results on native
+
+Command: 
+
+```
+RUSTFLAGS="-C target-cpu=native" cargo bench -- "2\^20"
+```
+
+### Sum of values
+
+```
+core_simd_sum 2^20 f32     [156.96 us 158.06 us 159.40 us]
+packed_simd_sum 2^20 f32   [184.17 us 184.47 us 184.85 us]
+nonsimd_sum 2^20 f32       [175.05 us 176.26 us 177.95 us]
+naive_sum 2^20 f32         [1.6636 ms 1.6700 ms 1.6778 ms]
+```
+
+### Sum of nullable values (`Vec<bool>`)
+
+```
+core_simd_sum null 2^20 f32   [2.3610 ms 2.3713 ms 2.3831 ms]
+packed_simd_sum null 2^20 f32 [1.5737 ms 1.5869 ms 1.6022 ms]
+nonsimd_sum null 2^20 f32     [1.8009 ms 1.8133 ms 1.8276 ms]
+naive_sum null 2^20 f32       [1.6418 ms 1.6520 ms 1.6660 ms]
+```
+
+### Sum of nullable values (`Bitmap`)
+
+```
+core_simd_sum bitmap 2^20 f32  [174.24 us 175.10 us 176.21 us]
+nonsimd_sum bitmap 2^20 f32    [541.78 us 545.16 us 549.09 us]
+naive_sum bitmap 2^20 f32      [1.6740 ms 1.6922 ms 1.7149 ms]
+```
+
+## Bench results on default
+
+Command: 
+
+```
+cargo bench -- "2\^20"
+```
 
 ### Sum of values
 
@@ -45,10 +85,36 @@ nonsimd_sum bitmap 2^20 f32    [454.78 us 462.08 us 471.82 us]
 naive_sum bitmap 2^20 f32      [1.7633 ms 1.7736 ms 1.7855 ms]
 ```
 
-### Conclusions so far:
+### Conditions
 
-* for non-null sums, it is advantageous (by 10%) to use `packed` or `core`
-* for sums with nulls, it is advantageous (by 50%) to use arrays
+```
+$ lscpu
+Architecture:        x86_64
+CPU op-mode(s):      32-bit, 64-bit
+Byte Order:          Little Endian
+CPU(s):              4
+On-line CPU(s) list: 0-3
+Thread(s) per core:  2
+Core(s) per socket:  2
+Socket(s):           1
+NUMA node(s):        1
+Vendor ID:           GenuineIntel
+CPU family:          6
+Model:               85
+Model name:          Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+Stepping:            4
+CPU MHz:             2095.077
+BogoMIPS:            4190.15
+Virtualization:      VT-x
+Hypervisor vendor:   Microsoft
+Virtualization type: full
+L1d cache:           32K
+L1i cache:           32K
+L2 cache:            1024K
+L3 cache:            36608K
+NUMA node0 CPU(s):   0-3
+Flags:               fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ss ht syscall nx pdpe1gb rdtscp lm constant_tsc rep_good nopl xtopology cpuid pni pclmulqdq vmx ssse3 fma cx16 pcid sse4_1 sse4_2 movbe popcnt aes xsave avx f16c rdrand hypervisor lahf_lm abm 3dnowprefetch invpcid_single pti tpr_shadow vnmi ept vpid fsgsbase bmi1 hle avx2 smep bmi2 erms invpcid rtm mpx avx512f avx512dq rdseed adx smap clflushopt avx512cd avx512bw avx512vl xsaveopt xsavec xsaves md_clear
+```
 
 ## License
 

benches/sum.rs

@@ -3,7 +3,7 @@ use criterion::{criterion_group, criterion_main, Criterion};
 use simd_benches::sum::*;
 
 fn close(l: f32, r: f32) {
-    assert!((l - r).abs() < l * 0.0001);
+    assert!((l - r).abs() < l * 0.001);
 }
 
 fn add_benchmark(c: &mut Criterion) {

benches/sum_nulls.rs

@@ -3,7 +3,7 @@ use criterion::{criterion_group, criterion_main, Criterion};
 use simd_benches::sum_nulls::*;
 
 fn close(l: f32, r: f32) {
-    assert!((l - r).abs() < l * 0.0001);
+    assert!((l - r).abs() < l * 0.001);
 }
 
 fn add_benchmark(c: &mut Criterion) {

benches/sum_nulls_bitmap.rs

@@ -4,7 +4,7 @@ use simd_benches::bitmap_ops;
 use simd_benches::sum_nulls_bitmap::*;
 
 fn close(l: f32, r: f32) {
-    assert!((l - r).abs() < l * 0.0001);
+    assert!((l - r).abs() < l * 0.001);
 }
 
 fn add_benchmark(c: &mut Criterion) {