Slower MySQL inserts for AMD64/Opteron?

[ray at redshift.com]

I think it's just a little odd that the AMD would be twice as fast with selects
yet 20% slower on inserts.  As far as variables, both machines are running
pretty much the same config.  I'm going to install i386 on the AMD and re-run
the tests to see if the problem has to do with something in i386 vs. AMD64.

Anyway, thanks.  Like you say, there are a few variables here - I was really
hoping to maybe run across someone that had been down this road before and could
comment on maybe something I'm over looking.  According to the FAE at AMD, even
on a bad day the Dual 246 opteron should drag the Xeon through the mud - but so
far nothing here has shown that to be the case.  It just feels like I've missed
some big thing I'm supposed to do to get the AMD in high gear on FreeBSD - but
so far I am at a loss as to what that is.  Nothing against AMD, but so far the
big solution (or potential solution) which has been suggested was to install
i386 so the AMD didn't have to deal with 64 bit pointer (or something along
those lines).  That's fine and I'm going to test that, but it just makes me
wonder why a possible solution would be to install the i386 code vs. the AMD64
code, when that's supposed to be much better right?

Anyway, I'm still hunting around, but I thought maybe someone might have some
suggestions and/or ideas to test out.

Ray


At 06:51 PM 6/22/2005 +0100, Steven Hartland wrote:
| Too many variables there too draw any conclusions from the comparison.
| Compare like for like.
| 
|     Steve / K
| ----- Original Message ----- 
| From: <ray at redshift.com>
| 
| 
| >I did some benchmarks recently on a single 2.4Ghz Xeon against a Dual 246
| > Opteron and noticed when it came to MySQL inserts, the single Xeon was about
20%
| > faster on inserts (and it actually had a slower hard drive, IDE vs SATA).  The
| > interesting thing is that the Dual Opterons was twice as fast retrieving the
| > data using selects.
| > 
| > Does anyone have any ideas as to why the single Xeon would beat it on inserts?
| > I've seen other benchmarks floating around (e.g. Tom's hardware guide) that
| > seems to mirror the results I found.  I also spoke to AMD about it and they
only
| > suggestion was to try running i386 on the AMD, as opposed to AMD64.  They also
| > mentioned something about perhaps compiler switches, but so far I haven't heard
| > back, nor have I had a chance to test further.
| > 
| > The Xeon was running 5.3/i386.  I believe the AMD was running 5.4/AMD64.
| > 
| > If anyone has any ideas as to what might be the cause, I would be interested in
| > hearing.  I'm getting ready to load i386 on the AMD (using a spare drive) just
| > to see if this makes a difference.  Am I over looking something that is
required
| > to unlock the power of the AMD?  The selects being twice as fast was great, but
| > then again the AMD has two CPU's also.
| 
| 
| ================================================
| This e.mail is private and confidential between Multiplay (UK) Ltd. and the
person or entity to whom it is addressed. In the event of misdirection, the
recipient is prohibited from using, copying, printing or otherwise disseminating
it or any information contained in it. 
| 
| In the event of misdirection, illegible or incomplete transmission please
telephone (023) 8024 3137
| or return the E.mail to postmaster at multiplay.co.uk.
| 
| 
|