MultiPI

Benchtype:

CPU and Memory

Memory footprint:

Low to very high

Description:

MultiPI use the Chudnovsky algorithm to compute digits of π, Link: Wikipedia.org

For now MaxxPI² can calculate up to 268.435.456 Digits of Pi (256M) and is designed as a multi-thread

function, all available CPU cores work together on a single calculation. It was made sure that all CPU's equally

served. As result, an intensive useage of MMX/SSEx and multithreading is given.

For further scaleability informations on different CPU types, see here: here.

MultiPI needs at least a dual-core processor and supports in the current version CPU's with 2,3,4,6,8 and 12 Core's.

This benchmark will only run on CPU's, with more than one core (HT core's counts as *real* core).

MultiPI is also used by MaxxPI-Preview-Multi.

Memory usage:

Detailed memory usage as follows (this may differ depending your used OS / amount cores):

(*) only available if x64 OS is present

Reached results (score):

Is given in K/sec., for easier compareing achieved results.

Instead of having to say 2min 32sec 343ms 123us, now you need only to tell one single number, K/sec.

It is nothing else than: the number of calculated decimal places per sec. in K (1024).

For example:

PI-calculation 1M became 356.2 K/sec. This means that the used CPU was able

to calculate 3.647.488 (1024 x 356.2) decimal places per second.

MaxxPI² also shows the used time for calculation (Calculated in) in the traditional way.

Further given informations:

Displaytime (Displayed in), shows the time needed to show the current resultset.

CRC (CRC32), shows the calculated crc32 value for the result

(it's used for validation-check (CRC Valid) with two results: NOT Valid! / Valid).

MaxxPI²