Skip to content
Permalink
master
Switch branches/tags

Name already in use

A tag already exists with the provided branch name. Many Git commands accept both tag and branch names, so creating this branch may cause unexpected behavior. Are you sure you want to create this branch?

MemoryBenchmarkWindows/mpe_examples/report_L1.txt

Go to file
 
 
Cannot retrieve contributors at this time
80 lines (69 sloc) 3.26 KB
Raw Blame
Open in GitHub Desktop
Memory Performance Engine. v0.60.06 for Windows x64. (C)2018 IC Book Labs.
get command line parameters...done
load API...done
detect processor features...done
detect SMP and cache levels...done
detect memory...done
detect paging...done
initializing domains builder...done
initializing threads builder...done
measure TSC clock...done
get library data...done
MPE native library. v0.22.00 for Windows x64. (C)2018 IC Book Labs.
run memory benchmark scenario.
asm method ID=12, bytes per instruction BPI=32
TSC frequency F=2404.256 MHz, period T=0.416 ns
cache points for data read/write:
L1=32 KB, L2=256 KB, L3=6144 KB, L4=0 KB
CPU core count=4, HyperThreading=1
memory allocation...done
NUMA-unaware mode, list allocated at base = 0000000000445640h
ID allocated base allocated size original (all hex)
---------------------------------------------------------
0 00000000001E0000 0000000000020000 00000000001E0000
---------------------------------------------------------
threads allocation...done
threads entries created = 1 , list allocated at base = 0000000000196F10h
ID (values hex)
-------------------------------------------------------------------------------
0 thread=00000000000000A0 rxevent=00000000000000E0 txevent=00000000000000E4
base1 =00000000001E0000 base2 =00000000001F0000 sizeins=0000000000000080
repeat=00000000001E8480 large page mode =0001 method id =000C
optimal affinity group\mask = 0000\0000000000000000
original affinity group\mask = 0000\0000000000000000
control set pointer = 0000000000445100
-------------------------------------------------------------------------------
running threads...done ( dTSC=00000000059E9D36h )
allocate statistics arrays...done
bandwidth statistics, BASE=0000000000C3A790h, SIZE=0000000000000080h
latency statistics, BASE=0000000000C3A820h, SIZE=0000000000000080h
RUN: method=12, threads=1, repeats=2000000
page mode=1, page size=4.00 KB
start=4.00 KB, end=64.00 KB, delta=4.00 KB, bpi=32 bytes
calibration...done ( delay=1.343 seconds, update repeats=1488664 )
[ read 256-bit AVX VMOVAPD ]
# size CPI nsPI MBPS
-------------------------------------------------------
1 4096 0.367 0.153 209774.359
2 8192 0.361 0.150 212884.343
3 12288 0.357 0.148 215747.225
4 16384 0.363 0.151 211910.093
5 20480 0.378 0.157 203765.496
6 24576 0.377 0.157 204228.925
7 28672 0.371 0.154 207379.497
8 32768 0.370 0.154 207781.853
9 36864 0.769 0.320 99983.668
10 40960 0.825 0.343 93268.977
11 45056 0.778 0.324 98917.746
12 49152 0.783 0.326 98225.786
13 53248 0.788 0.328 97674.547
14 57344 0.781 0.325 98508.618
15 61440 0.793 0.330 97043.207
16 65536 0.782 0.325 98441.416
-------------------------------------------------------
Bandwidth ( megabytes per second )
------------------------------------------------------------------------------
| max=215747.225 | min=93268.977 | average=153470.985 | median=151874.582 |
------------------------------------------------------------------------------
delete local objects...done
delete global objects...done