User Tools

Site Tools


linux:benchmark:cpu:pi

PI Berechnen

PI Berechnung auf 5000 Kommastellen. Dies lastet jedoch nur einen Core aus.

time echo "scale=5000; a(1)*4" | bc -l

Ergebnisse:

Hardware CPU Zeit
Selfmade Bender AMD x3 415e @ 2,5Ghz 37 Sec
Selfmade bastardo Via C6 @ 530Mhz 213 Sec
Dell Workstation Intel P4 @ 3Ghz 53 Sec
Lenovo X390 Intel i5-8265U @ 1,6Ghz 16 Sec
HP DL380 G8 Intel Xeon E5-2620 @ 2Ghz 28 Sec
Dell R7525 AMD Epyc 7452 @ 3,35Ghz 16 Sec
Raspberry Zero ARMv6 rev7 @ 700Mhz 146 Sec
Raspberry Pi 1 ??? 196 Sec
Raspberry Pi 3 ??? 67 Sec
Raspberry Pi 4 ARMv7 rev3 @ 2Ghz (OC) 37 Sec
  • Server bender 3x CPU AMD X3 415e @ 2,50Ghz: 37,5 Sec
  • Server bastardo 1x CPU VIA C6 (Nehemiah) @ 532Mhz: 213 Sec
  • Workstation trojar 1x CPU P4 @ 3Ghz: 53,9 Sec
  • Lenovo X390 i5-8265U CPU @ 1.60GHz: 16,5 Sec
  • HP DL380 G8 Xeon CPU E5-2620 0 @ 2.00GHz: 28 Sec
  • Dell R7525 AMD EPYC 7452 32-Core: 16 Sec
  • Raspberry 1: 3m 47 Sec
  • Raspberry 3: 1m 16 Sec
  • Raspberry 4: 37 Sec




time echo "scale=10000; a(1)*4" | bc -l

Ergebnisse:

Hardware CPU Zeit
Lenovo T460s Intel i7-6600U 37 Sec
Lenovo X390 Intel i5-8265U @ 1,6Ghz 94 Sec
HP DL380 G8 Intel Xeon E5-2620 @ 2Ghz 168 Sec
HP Microsurf G10 AMD Opteron X3431 2,10Ghz 157 Sec
Dell R7525 AMD Epyc 7452 @ 3,35Ghz 96 Sec



PI Multi Start

for i in $(seq <cpucores>)  ; do echo "scale=5000; a(1)*4" | bc -l & done

Beispiel für 32 Cores:
for i in $(seq 32)  ; do echo "scale=5000; a(1)*4" | bc -l & done
linux/benchmark/cpu/pi.txt · Last modified: by stone