At 07:41 PM 3/22/2006, [EMAIL PROTECTED] wrote:
Dear Great Minds
Not strictly a Beowulf question but I suspect knowledge on the topic lurks
herein.
I can't seem to find a straight answer on the "real" *minimum* cable
length for
gigabit ethernet. I'm looking to make things on my new toy nice
Dear Great Minds
Not strictly a Beowulf question but I suspect knowledge on the topic lurks
herein.
I can't seem to find a straight answer on the "real" *minimum* cable length for
gigabit ethernet. I'm looking to make things on my new toy nice and tidy.
Assume Cat5e/Cat6 cable, Intel e1000 NIC
At 11:16 AM 3/22/2006, Vincent Diepeveen wrote:
- Original Message - From: "Jim Lux" <[EMAIL PROTECTED]>
If it's a full custom chip, figure a "first chip" cost of $2M. (layout, a
couple spins, etc., but assuming you know basically what the chip is
supposed to do and how to do it)
M
> but something intended to cream off the university world. Those universities
waste money by the shitload,
[Troll]
Yeah, what have the Romans^H^H^H^H^H^HUniversities ever done for us? Look at the
commercial world for how things *should* be done... companies like MCI Worldcomm
or Enron.
And as f
On Wed, 22 Mar 2006, Lawrence Stewart wrote:
> I'll just be attending to look at the exhibits because I am local,
> but it would be great to say
As usual, I'll be there.
And as usual, we have a new release ready in time for LinuxWorld. But we
aren't doing any big technology introductions (th
I'll just be attending to look at the exhibits because I am local,
but it would be great to say
hello.
-Larry Stewart / engineer at SiCortex building a cool, but still
stealthy, cluster machine
___
Beowulf mailing list, Beowulf@beowulf.org
To cha
Should the question not be: how much are you willing to pay. If you only
get a 2X speed up in your run time for $10K, then you will buy more nodes.
If you get a 10X, you may call it a wash and could go either way. If you
see a 1000X then you will pay a great deal more than $10K. I have seen a
bi
- Original Message -
From: "Bill Harman" <[EMAIL PROTECTED]>
To: "'Vincent Diepeveen'" <[EMAIL PROTECTED]>; <[EMAIL PROTECTED]>;
"'Jim Lux'" <[EMAIL PROTECTED]>
Cc:
Sent: Wednesday, March 22, 2006 11:05 PM
Subject: RE: [Beowulf] Vector coprocessors AND CILK
Should the question not
At 01:24 PM 3/22/2006, David Mathog wrote:
Anybody know how many g's a typical disk can withstand before it fails?
A lot more than 1g...
You can take a (running) disk and move it through a variety of orientations
without any problem. That would put a lot more load on the bearings, etc.
than
Anybody know how many g's a typical disk can withstand before it fails?
My earthquake bolted rack might stay upright through the big 7.0
and yet still be very much damaged afterwards due to head crashes
on the disks.
Thanks,
David Mathog
[EMAIL PROTECTED]
Manager, Sequence Analysis Facility, Bio
On Wed, 2006-03-22 at 09:00 -0500, Luis Alejandro del Castillo Riley
wrote:
> Hi fellows i have 5 pcs one is my master computer and the others are
> the slaves machines i am trying to build a cluster and i am looking
> for a program can do a remote console
> or desktop from the master pc to the oth
I'm
not sure if this is what you are looking for, but www.realvnc.com allows you to remote control
across different platforms. On most distros of Linux, a VNC server is
already included (x11vnc is the server included with xwindows, krdc is a client
included with kde). Most (all?) of these
- Original Message -
From: "Jim Lux" <[EMAIL PROTECTED]>
To: "Vincent Diepeveen" <[EMAIL PROTECTED]>; <[EMAIL PROTECTED]>
Cc:
Sent: Wednesday, March 22, 2006 6:00 AM
Subject: Re: [Beowulf] Vector coprocessors AND CILK
At 07:18 PM 3/21/2006, Vincent Diepeveen wrote:
- Original M
*** Call for Papers and Announcement ***
Wavelet Applications in Industrial Processing IV (SA209)
Part of SPIE’s International Symposium on Optics East 2006
1-4 October 2006 • Boston Marriott Copley Place • Boston, MA, USA
--- Abstract Due Date: 25 March 2006 ---
--- Manuscript Due Date: 4 Sep
Hi fellows i have 5 pcs one is my master computer and the others are
the slaves machines i am trying to build a cluster and i am looking for
a program can do a remote console
or desktop from the master pc to the others but not using SSH.
The whole idea is find a remplacement of a KVM switch.
__
Title: RS: [Beowulf] Cluster newbie, power recommendations
I have done this for several years. If you remove the video card, CD/DVD units and other power consumers, leaving just a single hard disk and network, a modern PSU in this range is well capable of handling at least two motherboards.
> Why bother with virtualization? You can do this with any of the freely
> available MPI versions.
Sure, or PVM.
But there's more to running a cluster than just running the
parallel code. With virtualization it would be possible to
practice setting up a cluster as well, without actually
need
At 06:27 AM 3/22/2006, Andrew Piskorski wrote:
On
So far the double sided foam tape seems to work very well! I've tried
it on a SMALL scale - 3 motherboards. They run memtest86+ just fine,
but I have not yet tried network booting Linux on them with Warewulf
and actually using them heavily in a
Here is the latest on Cluster Monkey:
- MPI Debugging (part 1)
- Parallel Programming
- A new Poll! Cluster Books
- Essential Cluster News
http://www.clustermonkey.net
No cookie trays, but we were thinking of mounting motherboards
on the outside of a car so we could cool them by driving ar
On Tue, Mar 21, 2006 at 06:58:05AM -0500, Robert G. Brown wrote:
Subject: Re: [Beowulf] Cluster newbie, power recommendations
> On Mon, 20 Mar 2006, Charlie Peck wrote:
> >signal back to the power supply. I believe the site that RGB is referring
> >to is http://joule.bu.edu/~hazen/LinuxCluster.
On Sun, Mar 19, 2006 at 07:09:42PM -0600, Eric Geater at Home wrote:
> My greatest question, though, revolves around power distribution. It seems
> kinda weak to simply use one PSU per motherboard, especially if I take the
> time to devise a cabinet in which to operate the goodies. Is it possibl
On Tue, Mar 21, 2006 at 10:32:30PM +, John Hearns wrote:
> > Not sure of the performance impact of this, but you could look at OpenVZ
> > or Xen as well (when it is ready).
> Xen has very little impact on performance. I saw some very good figures
> at a recent presentation at FOSDEM.
The v
22 matches
Mail list logo
