MathWorld Headline News
42nd Mersenne Prime Found
By Eric W. Weisstein
February 26, 2005--Less than a year after the 41st Mersenne prime was reported (MathWorld headline news: June 1, 2004), the Great Internet Mersenne Prime Search (GIMPS) project has discovered the 42nd known Mersenne prime. The candidate prime was flagged prime by Dr. Martin Nowak on February 18, independently verified by Tony Reix on February 25, and the exponent was reported on February 26 (Reix). Nowak's calculation required more than 50 days on a 2.4-GHz Pentium 4 computer, while verification took 5 days on a 16 Itanium CPU Bull NovaScale 5000 HPC. Additional details can be found in the Mersenne.org press release.
Mersenne numbers are numbers of the form Mn = 2n - 1, giving the first few as 1, 3, 7, 15, 31, 63, 127, .... Interestingly, the definition of these numbers therefore means that the nth Mersenne number is simply a string of n 1s when represented in binary. For example, M7 = 27 - 1 = 127 = 11111112 is a Mersenne number. Mersenne primes are Mersenne numbers that are also prime, i.e., have no factors other than 1 and themselves. So, since the number 127 is prime and is a Mersenne number, it is a Mersenne prime.
The new Mersenne prime is 225,964,951 - 1 = 12216463006127794810...98933257280577077247 (where the ellipses indicate that several million intervening digits have being omitted for conciseness) and has a whopping total of 7,816,230 decimal digits. It is therefore not only the largest known Mersenne prime, but also the largest known prime of any kind. In fact, there is a particularly efficient and, more importantly, deterministic primality test for Mersenne numbers known as the Lucas-Lehmer test. The efficiency of this test combined with the high historical profile of the Mersenne numbers thus accounts for the fact that the four largest known primes are all Mersenne primes.
For those curious to see the new prime in its full 7,816,230 digits of glory, the results of a short Mathematica calculation generating its decimal digits are available by downloading the notebook mersenne42.nb. If you do not own Mathematica, you can download a free trial version to view this file. A poster featuring all 7.8 million digits of the new prime (displayed in an extremely small point size) created by Richard Crandall, discoverer of the advanced transform algorithm used by the GIMPS program, is (or will shortly be) available from Perfectly Scientific.
The eight largest known Mersenne primes (including the latest) have all been discovered by GIMPS, which is a distributed computing project being undertaken by an international collaboration of volunteers. As of February 23, 2005, GIMPS participants have tested and double-checked all exponents n below 9,889,900, while all exponents below 15,130,000 have been tested at least once.
The study of such numbers has a long and interesting history, and the search for Mersenne numbers that are prime has been a computationally challenging exercise requiring the world's fastest computers. Mersenne primes are intimately connected with so-called perfect numbers, which were extensively studied by the ancient Greeks, including by Euclid. A complete list of indices n of the previously known Mersenne primes is given in the table below (as well as by sequence A000043 in Neil Sloane's On-Line Encyclopedia of Integer Sequences). However, note that the region between the 39th and 40th known Mersenne primes has not been completely searched, so while the 40th number listed is the 40th Mersenne prime discovered, it is not yet known if M20,996,011 is actually the 40th Mersenne prime.
# | n | digits | year | discoverer (reference) |
1 | 2 | 1 | antiquity | |
2 | 3 | 1 | antiquity | |
3 | 5 | 2 | antiquity | |
4 | 7 | 3 | antiquity | |
5 | 13 | 4 | 1461 | Reguis (1536), Cataldi (1603) |
6 | 17 | 6 | 1588 | Cataldi (1603) |
7 | 19 | 6 | 1588 | Cataldi (1603) |
8 | 31 | 10 | 1750 | Euler (1772) |
9 | 61 | 19 | 1883 | Pervouchine (1883), Seelhoff (1886) |
10 | 89 | 27 | 1911 | Powers (1911) |
11 | 107 | 33 | 1913 | Powers (1914) |
12 | 127 | 39 | 1876 | Lucas (1876) |
13 | 521 | 157 | Jan. 30, 1952 | Robinson |
14 | 607 | 183 | Jan. 30, 1952 | Robinson |
15 | 1279 | 386 | Jan. 30, 1952 | Robinson |
16 | 2203 | 664 | Jan. 30, 1952 | Robinson |
17 | 2281 | 687 | Jan. 30, 1952 | Robinson |
18 | 3217 | 969 | Sep. 8, 1957 | Riesel |
19 | 4253 | 1281 | Nov. 3, 1961 | Hurwitz |
20 | 4423 | 1332 | Nov. 3, 1961 | Hurwitz |
21 | 9689 | 2917 | May 11, 1963 | Gillies (1964) |
22 | 9941 | 2993 | May 16, 1963 | Gillies (1964) |
23 | 11213 | 3376 | Jun. 2, 1963 | Gillies (1964) |
24 | 19937 | 6002 | Mar. 4, 1971 | Tuckerman (1971) |
25 | 21701 | 6533 | Oct. 30, 1978 | Noll and Nickel (1980) |
26 | 23209 | 6987 | Feb. 9, 1979 | Noll (Noll and Nickel 1980) |
27 | 44497 | 13395 | Apr. 8, 1979 | Nelson and Slowinski (Slowinski 1978-79) |
28 | 86243 | 25962 | Sep. 25, 1982 | Slowinski |
29 | 110503 | 33265 | Jan. 28, 1988 | Colquitt and Welsh (1991) |
30 | 132049 | 39751 | Sep. 20, 1983 | Slowinski |
31 | 216091 | 65050 | Sep. 6, 1985 | Slowinski |
32 | 756839 | 227832 | Feb. 19, 1992 | Slowinski and Gage |
33 | 859433 | 258716 | Jan. 10, 1994 | Slowinski and Gage |
34 | 1257787 | 378632 | Sep. 3, 1996 | Slowinski and Gage |
35 | 1398269 | 420921 | Nov. 12, 1996 | Joel Armengaud/GIMPS |
36 | 2976221 | 895832 | Aug. 24, 1997 | Gordon Spence/GIMPS (Devlin 1997) |
37 | 3021377 | 909526 | Jan. 27, 1998 | Roland Clarkson/GIMPS |
38 | 6972593 | 2098960 | Jun. 1, 1999 | Nayan Hajratwala/GIMPS |
39 | 13466917 | 4053946 | Nov. 14, 2001 | Michael Cameron/GIMPS (Whitehouse 2001, Weisstein 2001) |
40? | 20996011 | 6320430 | Nov. 17, 2003 | Michael Shafer/GIMPS (Weisstein 2003) |
41? | 24036583 | 7235733 | May 15, 2004 | Josh Findley/GIMPS (Weisstein 2004) |
42? | 25964951 | 7816230 | Feb. 18, 2005 | Martin Nowak/GIMPS (Weisstein 2005) |
Caldwell, C. K. "The Largest Known Primes." http://www.utm.edu/research/primes/largest.html
GIMPS: The Great Internet Mersenne Prime Search. http://www.mersenne.org
GIMPS: The Great Internet Mersenne Prime Search Status. http://www.mersenne.org/status.htm
Mersenne.org. "Mersenne.org Project Discovers New Largest Known Prime Number, 225,964,951 - 1." Feb. 27, 2005. http://www.mersenne.org/25964951.htm
Reix, T. "The GIMPS Project has Found a new Mersenne Prime Number: M42." Message to NMBRTHRY@LISTSERV.NODAK.EDU. Feb. 26, 2005. http://listserv.nodak.edu/cgi-bin/wa.exe?A2=ind0502&L=nmbrthry&F=&S=&P=2115
Weisstein, E. W. "MathWorld Headline News: 40th Mersenne Prime Announced." Dec. 2, 2003. http://mathworld.wolfram.com/news/2003-12-02/mersenne
Weisstein, E. W. "MathWorld Headline News: 41st Mersenne Prime Announced." Jun. 1, 2004. http://mathworld.wolfram.com/news/2004-06-01/mersenne
Weisstein, E. W. "MathWorld Headline News: 42nd Mersenne Prime (Probably) Discovered." Feb. 18, 2005. http://mathworld.wolfram.com/news/2005-02-18/mersenne
Weisstein, E. W. "The 42nd Mersenne Prime." http://mathworld.wolfram.com/news/2005-02-18/mersenne/mernsenne42.nb
Woltman, G. "New Mersenne Prime?!" Message to The Great Internet Mersenne Prime Search List. Feb. 18, 2005.
Woltman, G. "42nd Mersenne Prime." Message to The Great Internet Mersenne Prime Search List. Feb. 25, 2005.