How Fast Can You Crunch Pi to a billion decimals?

279 seconds for me (before verification). Core i7 4770, 16GB RAM. Running the x64 AVX - Hina executable optimized for Sandy Bridge and later processors.

Mmmmmm. Crunch Pie!

i5 2500K / 8GB RAM.

From SSD (SATA 3 GB/s Agility 2) - 315 Seconds.
From HDD (5400RPM SATA) - 347 Seconds.

Yea, I/O bottlenecked.

(That was just the version off the main page…is there a sub-page with different versions?)

If only this forum had a ‘like’ function.

By hand?

i5 3570k @3.4Ghz, 16GB RAM, 303.168 seconds (all in RAM). Though I confess I have little idea what most of the options mean.

Ah, computing speed and number crunching, good fun!

Total Computation Time:             528.750 seconds  ( 0.147 hours )
Total Time (with output + verify):  559.089 seconds  ( 0.155 hours )

CPU Utilization:        753.401 %
Multi-core Efficiency:  94.1751 %

Last Digits:  Pi
6434543524 2766553567 4357021939 6394581990 5483278746  :  999,999,950
7139868209 3196353628 2046127557 1517139511 5275045519  :  1,000,000,000

Version:          0.6.4 Build 9424 (x64 SSE4.1 - Windows ~ Ushio)
Processor(s):     Intel(R) Core(TM) i7 CPU 920 @ 2.67GHz
Logical Cores:    8
Physical Memory:  12,875,452,416 ( 12.0 GB )
CPU Frequency:    2,673,336,624 Hz  (frequency may be inaccurate)

I think the main download has all the different versions…the binaries folder of my download has 10 different executables. Running y-cruncher.exe picked one for me (x64 SSE4.1 ~ Ushio).

Ah right :)

Yea, it picked the Hina AVX one for me.

My computer doesn’t meet the minimum requirements to run the executable.

Did I do this right? Here are my results, it looks like 200.505 seconds before verification? I have never heard my CPU fan kick in before, the temp (which has never gone above 50) hit 88 or so during this test. Probably good to put my new rig through its paces, but my heart was in my throat the entire time.

I did a run while watching the temps. Towards the end my CPU was up to 61-62C, which is only 1-2 degrees off what it topped at during a long Prime95 run. I’d still use Prime95 for stress-testing, but this program will give your CPU a good workout.

Processor(s):          Intel(R) Core(TM) i5-4670 CPU @ 3.40GHz
Logical Cores:         4
Physical Memory:       17,096,310,784 bytes  ( 16.0 GiB )
CPU Frequency:         3,399,082,031 Hz

Program Version:       0.6.5 Build 9442 (x64 AVX2 - Windows ~ Airi)
Constant:              Pi
Algorithm:             Chudnovsky Formula
Decimal Digits:        1,000,000,000
Hexadecimal Digits:    Disabled
Threading Mode:        4 threads
Computation Mode:      Ram Only
Working Memory:        4,860,298,856 bytes  ( 4.52 GiB )
Logical Disk Usage:    0 bytes  ( 0 bytes )

Start Date:            Tue May 27 21:34:19 2014
End Date:              Tue May 27 21:39:10 2014

Computation Time:      276.567 seconds
Total Time:            290.569 seconds

CPU Utilization:           389.717 %
Multi-core Efficiency:     97.429 %

Last Digits:
6434543524 2766553567 4357021939 6394581990 5483278746  :  999,999,950
7139868209 3196353628 2046127557 1517139511 5275045519  :  1,000,000,000

I don’t even know if that’s a good speed for an i5. Most of the times on the benchmarks page are from i7 and Xeon CPUs.

Interesting, they got an AVX2 version working then… (That’s Haswell only atm)

Seems to be /slightly/ faster than a direct comparison with my i5 2500K would be.

I get 312s on my i5-2500K, so 3s off last build (on a different compiler for this version, so I checked :) )

294 seconds for me.

Boy, I knew my 4-year-old Intel i7 920 was getting long in the tooth (in computing hardware years, anyway) but this kinda drives it home. 40% or more slower than most of the rest of your results. Fortunately, almost nothing I do is CPU-bound so the old dog still handles his workload just fine.