http://sqlblog.com/blogs/paul_white/archive/2011/08/09/sql-server-seeks-and-binary-search.aspxThe following table summarizes the results from my last two blog entries, showing the CPU time used when performing 5 million clustered index seeks: In test 1, making the clustered index unique improved performance by around 40%. In test 2, making theenCommunityServer 2.1 SP2 (Build: 61129.1)http://sqlblog.com/blogs/paul_white/archive/2011/08/09/sql-server-seeks-and-binary-search.aspx#37625Mon, 08 Aug 2011 21:11:00 GMT21093a07-8b3d-42db-8cbf-3350fcbf5496:37625Rob Farley<p>Nice, Paul.</p> http://sqlblog.com/blogs/paul_white/archive/2011/08/09/sql-server-seeks-and-binary-search.aspx#37626Mon, 08 Aug 2011 21:20:46 GMT21093a07-8b3d-42db-8cbf-3350fcbf5496:37626Paul White<p>Cheers, Rob.</p> http://sqlblog.com/blogs/paul_white/archive/2011/08/09/sql-server-seeks-and-binary-search.aspx#37627Mon, 08 Aug 2011 21:23:23 GMT21093a07-8b3d-42db-8cbf-3350fcbf5496:37627Dylan<p>Fascinating read as always!</p>http://sqlblog.com/blogs/paul_white/archive/2011/08/09/sql-server-seeks-and-binary-search.aspx#37629Tue, 09 Aug 2011 04:00:00 GMT21093a07-8b3d-42db-8cbf-3350fcbf5496:37629Alexander Kuznetsov<p>Paul,</p> <p>This is a brilliant post, of course, but I am not completely with you on the following: &quot;Linear interpolation has the potential to be more efficient even than binary search&quot; - there are some exceptions when a slow but reliable binary search converges much faster than its smarter alternatives.</p> http://sqlblog.com/blogs/paul_white/archive/2011/08/09/sql-server-seeks-and-binary-search.aspx#37630Tue, 09 Aug 2011 07:17:21 GMT21093a07-8b3d-42db-8cbf-3350fcbf5496:37630csm<p>Great. It's always a pleasure read articles like this.</p>http://sqlblog.com/blogs/paul_white/archive/2011/08/09/sql-server-seeks-and-binary-search.aspx#37637Tue, 09 Aug 2011 11:27:04 GMT21093a07-8b3d-42db-8cbf-3350fcbf5496:37637tobi<p>Who would have guessed this explanation?</p>http://sqlblog.com/blogs/paul_white/archive/2011/08/09/sql-server-seeks-and-binary-search.aspx#37641Tue, 09 Aug 2011 15:50:02 GMT21093a07-8b3d-42db-8cbf-3350fcbf5496:37641Paul White<p>@Alex: Agreed, and that's why I said &quot;has the potential&quot;. &nbsp;There is much more detail in the papers referenced at the end of the post.</p> <p>@csm: Thank you</p> <p>@tobi: I didn't hear from anyone that guessed correctly. &nbsp;Perhaps I should have offered a prize ;c)</p> http://sqlblog.com/blogs/paul_white/archive/2011/08/09/sql-server-seeks-and-binary-search.aspx#45101Mon, 10 Sep 2012 06:59:22 GMT21093a07-8b3d-42db-8cbf-3350fcbf5496:45101amit<p>its 2 gud dude................</p>http://sqlblog.com/blogs/paul_white/archive/2011/08/09/sql-server-seeks-and-binary-search.aspx#48749Wed, 17 Apr 2013 07:24:56 GMT21093a07-8b3d-42db-8cbf-3350fcbf5496:48749Boris Hristov<p>Insanely great article, Paul! </p>