53位的回文平方数

[FengYuanMSFT]

Six years ago I wrote a program to search for palindromic square numbers and other palindromic numbers and found some new palindromic numbers. But with increasing length, the search became longer and longer so I finally gave it up.

Recently I got a new powerful machine with dual quad-core CPUs, so I brought it home during the Christmas and New Year holiday season. I modified my old search program to take advantage of multi-cores in the machine. The search program has tons of parallelism inside. Although the original code was written as incremental loops for performance reason, it is not hard to change it to run on multi-core system. The result is almost 8 fold improvement in speed, on top of CPU speed gain for the last six years.

With the multi-core version, I tried continuing the old quest again and find quite a few record breaking palindromic numbers during the holiday season. But I did not try to break the 52-digit palindromic square number record held by Pete Leadbetter since May 20, 2001 until last night. Before leaving work, I started searching for 53 digit paradromic sqaures and this morning we have a new record, a 53-digit palindromic square number:

122063831551139898460740721 ^ 2 = 14899578972945056149893218681239894165054927987599841

Other palindromic number records can be found at:

http://www.worldofnumbers.com/palrecs.htm

Go MultiCore!

[Chivalry]

ff

[sanguniang]

很难，很深奥！

[jspadmin]

Faint
-_-

[jspadmin]

楼主是想告诉我们：其实双核的机器，更适合去寻找更多位数的回文平方数。

P.S.:回文数是指以一个数的中间一个数字为对称中心，两边的数字是中心对称的，比如131，12321。回文平方数，是刚好这数又是另一个整数的平方结果。

[brightyang]

faint

[alen88]

看看

[guanbenben]

是这样

[AFIC]

纯灌水，贴到算法版，没准5分钟后有人就贴出54位回文平方数也说不定……
虽说没啥实际意义吧。

[FreezingseaZeng]

盲目地顶老大的贴~

[bigpeon]

Orz

[ouyh12345]

up

[FengYuanMSFT]

在 专题开发/技术/项目数据结构与算法 发了一贴:

http://topic.csdn.net/u/20080117/15/e97a8d4d-f1cd-4e65-b46b-dd4352fe4c0a.html


回文平方数的记录原来是 Pete Leadbetter May 20, 2001 的:

64897400105515621177314682^2 = 4211672540455378958718869999688178598735540452761124 (52位).

我昨天找到了一个 53 位的回文平方数:

122063831551139898460740721^2 = 14899578972945056149893218681239894165054927987599841

用的机器是 DELL PRECISION 9200 DUAL QUADCORE 2.33 GHZ, C++, 算法用了八个 CORE. 不过还没算完所有 52 位数, 所以八个 CORE对找到这个数作用不大.

其他的一些回文数记录在:

http://www.worldofnumbers.com/palrecs.htm

用兴趣的可以试试. 不过简单的不算: 1 0000000 .... 1 ^2 很多是回文数, 1 000099..000 .... 1 ^2 也不少.

[ToperRay]

仰望中。。。

[一条晚起的虫]

只是说明现在机器可以支持以前无法支持的算法了。

[shenglijay]

多核.. go go go

[传说中的堕落]

看个热闹

[ydone]

顶袁老大的，俺目前还没有接触到双核的pc机，很落后。

[michaelwangwh]

顶袁老大的，为什么双核的机器能更有效率的找平方回文呢？

[gqcwin]

uP

[m60a1]

晕呼，双核的PC机，如果你用二种不同牌子双核CPU试一下，你还会发现有意思的现象!
(我所指的是只让此PC做纯计算工作！)