prime95のエラー

メルセンヌ素数を発見する分散コンピューティングプロジェクトのGIMPS、相変わらず参加している。ずっと起動しているprime95をふと確認してみるとエラーメッセージを吐き続けている。次のような感じで延々と。

[Jan 24 22:33] Iteration: 50800000 / 54082339 [93.93%]. Per iteration time: 0.057 sec.
[Jan 24 22:33] Possible hardware errors have occurred during the test! 1 ROUNDOFF > 0.4.
[Jan 24 22:33] Confidence in final result is fair.
[Jan 24 22:43] Iteration: 50810000 / 54082339 [93.94%]. Per iteration time: 0.057 sec.
[Jan 24 22:43] Possible hardware errors have occurred during the test! 1 ROUNDOFF > 0.4.
[Jan 24 22:43] Confidence in final result is fair.
[Jan 24 22:52] Iteration: 50820000 / 54082339 [93.96%]. Per iteration time: 0.057 sec.
[Jan 24 22:52] Possible hardware errors have occurred during the test! 1 ROUNDOFF > 0.4.
[Jan 24 22:52] Confidence in final result is fair.
[Jan 24 23:02] Iteration: 50830000 / 54082339 [93.98%]. Per iteration time: 0.057 sec.
[Jan 24 23:02] Possible hardware errors have occurred during the test! 1 ROUNDOFF > 0.4.
[Jan 24 23:02] Confidence in final result is fair.
[Jan 24 23:11] Iteration: 50840000 / 54082339 [94.00%]. Per iteration time: 0.057 sec.
[Jan 24 23:11] Possible hardware errors have occurred during the test! 1 ROUNDOFF > 0.4.
[Jan 24 23:11] Confidence in final result is fair.
[Jan 24 23:21] Iteration: 50850000 / 54082339 [94.02%]. Per iteration time: 0.057 sec.
[Jan 24 23:21] Possible hardware errors have occurred during the test! 1 ROUNDOFF > 0.4.
[Jan 24 23:21] Confidence in final result is fair.

また、result.txtを開いてみると、次の一行が見つかった。

Iteration: 40025344/54082339, ERROR: ROUND OFF (0.5) > 0.40

なんのこっちゃということで調べてみると、フォーラムに質問が出ていた。エラーが1回でも起こると、そのナンバーが終わるまで、スクリーンに延々とメッセージを吐き続ける仕様らしい。連発してエラーが起こっているのかと思ったが、そうではないということで一安心である。