Doesn't work properly on Huawei Ascend P1 Calculates 1000 digits ok, but when calculating 1000000 the results displayed are garbage numbers, and not the same every time. They might be digits at a random position in the middle of Pi if the calculation succeeded. Also "view externally" options don't work. Good attempt though porting a parallel Pi algorithm to Android. Note that the tail end of the calculation is not parallelized, occupies only one core on my phone.
Good CPU Stability Test. Update: Scored 38.7 on Nexus 4 oc 1.6Ghz. Franco kernel 142 AOKP 4.2.2. A tip; make sure you device is cool before running any form of benchmarks, thermal throttling will prevent high scores if the device gets too hot. I scored 47 secs initially when by device was scorching hot..
Slow Take the performance of a Surface Pro 4 (i5-6300U) which is on par with Galaxy S7 (the real version - Exynos) and it can do 10M in half the time that this takes with Mathematica, on a single core. And Mathematica isn't even optimised for Pi. Learn some high level mathematics, and maybe some assembly while you are at it. ARM should have a performance edge over X86 because the instructions allow putting the result into a different register.
19.7 to 10million. #1. From Galaxy Note 5. I love this phone. Lightning fast at everything. And....this fabulous SCREEN!!! Best phone I've ever owned and I've had many. The pictures this thing takes too..oh Lord. My wife's iPhone 6 pales in comparison.
Bad results 5 sec for 1 000 000?! Something is wrong with this metric, it seems to be imposible Thanks for respond now 68 sec with RK3066 tablet (2x1.6GHz/1MB L2, ARMv7 NEON)
Why my Samsung S3 shows 113s while author says 46s? Other of my data make sense: Shield portable-27s; LG volt 69s; Moto E 2nd Gen: 51s
- New ranking page to see how fast of the device
- Rotation fixed to no-sensor
- Fix several possible bugs
- Add internal viewer for Pi's first 100,000 digits
- You can still view full digits in external text viewer
- Fix several crashes