Statalist


[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

Re: st: re: Memory, Stata 10, and 32-bit XP


From   "Fred Wolfe" <[email protected]>
To   [email protected]
Subject   Re: st: re: Memory, Stata 10, and 32-bit XP
Date   Sat, 23 Feb 2008 09:43:34 -0600

Thanks to everyone for their comments. Kit is correct, I meant to say
Mac Pro, not Power Mac.

Thanks to Steve for his reference. Another reference of use,
particularly with respect to the Mac Pro, is
http://www.barefeats.com/harper9.html .It shows the advantage of Raid
0 for large sequential reads and writes, but no advantage for smaller
reads and writes.

When we compared Aa vista (2 core) with an 8 core (4 core Stata
enabled) Mac Pro, a series of programs we run nightly ran in 6 hours
compared to 12 hours. We were very impressed. We take a series of SQL
files, use StatTransfer to move them to Stata, and then build and test
a series of Stata analytic files (~750MB). Because these program
involve merging, preserving, restoring, reading and saving many files
repeatedly, we were interested in whether we could achieve additional
time gains by using Raid structures. Based on reading, but not on data
examination, we would guess that the benefits of Raid would be small.
So, right now it looks like no Raid for us.

Thanks everyone who commented.

Fred

On 2/22/08, Steven Dubnoff <[email protected]> wrote:
> Fred wrote:
>
>   > We are wondering if Kit (and others) have
>  >experience with the extent to which performance is boosted (+ or -0)
>  >by Raid 0, 1 or 5) using Stata, and if there are any recommendations.
>
>
> In general, Raid 0 (and 5) are highly overrated and only will
>  increase performance in heavily loaded server environments.  Raid 0
>  doubles the probability of failure, since the failure of one disk
>  makes the data on both drives unusable. See this:
>
>  http://faq.storagereview.com/tiki-index.php?page=SingleDriveVsRaid0
>
>  Raid 1 is good for reliability and has minimal impact on
>  performance.  Since, as Kit pointed out, Stata hardly ever hits your
>  disks, it will not matter.  The same is not true for SAS, SPSS and
>  other disk-based systems.
>
>  Best,
>
>  Steve
>
>
>
>  ----------------------------------------------------------------
>
> Steven Dubnoff              [email protected]
>  Circle Systems              We make your data instantly usable.
>  Download Stat/Transfer from http://www.stattransfer.com
>  1001 Fourth Avenue, #3200   (206) 682-3783
>  Seattle, WA 98154           Fax (206) 328-4788
>
>
>
>  *
>  *   For searches and help try:
>  *   http://www.stata.com/support/faqs/res/findit.html
>  *   http://www.stata.com/support/statalist/faq
>  *   http://www.ats.ucla.edu/stat/stata/
>


-- 
Fred Wolfe
National Data Bank for Rheumatic Diseases
Wichita, Kansas
NDB Office  +1 316 263 2125 Ext 0
Research Office +1 316 686 9195
[email protected]
*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/



© Copyright 1996–2024 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index