[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: new backend computer? (fwd)
- To: Peter Kammel <kammel@npl.uiuc.edu>
- Subject: Re: new backend computer? (fwd)
- From: Frederick Gray <fegray@berkeley.edu>
- Date: Fri, 27 Jan 2006 12:37:36 -0800
- Cc: dwebber@npl.uiuc.edu
- In-reply-to: <Pine.LNX.4.63.0601271400590.4076@one.npl.uiuc.edu>
- References: <Pine.LNX.4.63.0601271400590.4076@one.npl.uiuc.edu>
- User-agent: Mutt/1.5.9i
On Fri, Jan 27, 2006 at 02:04:12PM -0600, Peter Kammel wrote:
> Dear Fred,
>
> FYI. Everybody agrees that your idea is good. Opinions differ
> a bit whether we get the most for our money with the
> $3400 machine. What is the exact configuration you
> are envisioning.
>
Hi, Peter,
The "$3400 machine" was intended as a starting point for discussion, and I
would also hope that we can optimize the price/performance ratio. I had been
looking at the dual Opteron 270 machine from the vendor that sold us cheetah:
http://www.swt.com/dualo5.html .
With a 250 GB drive rather than the default 80 GB, the current price is
$3271 plus shipping.
One of my KamLAND colleagues just bought an online processing farm of
Dell PowerEdge 1850 dual Xeon 3.2 GHz machines. I have not benchmarked
our software on a Xeon, so I don't know how to compare this configuration
directly. I suspect that it is somewhat slower. With 2 * 146 GB SCSI hard
drives, it currently prices out at $3023.70.
Also, one reply to David's comment:
>Would a faster backend machine be a good investment? MuCap is considering
>purchasing a new machine for analysis that would be four times faster than
>their Cheetah, [...]
Actually only two times. The other factor of 2 comes from using both CPUs for
analysis, rather than one for analysis and one for event building and
other tasks.
Thanks,
-- Fred
-- Fred Gray / Visiting Postdoctoral Researcher --
-- Department of Physics / University of California, Berkeley --
-- fegray@berkeley.edu / phone 510-642-2438 / fax 510-642-9811 --