Re: [Beowulf] Advantage to compiling RHEL 5 with opteron option?

2007-09-26 Thread Chris Samuel
On Thu, 27 Sep 2007, Jeremy Fleming wrote: > Anyone know if there is any advantage to recompiling the default > RHEL 5 kernel to include the "Opteron" config option on a quad > processor opteron machine? It may well be worth your while moving to the current kernel from the RHEL one (2.6.22.9), w

[Beowulf] Advantage to compiling RHEL 5 with opteron option?

2007-09-26 Thread Jeremy Fleming
Anyone know if there is any advantage to recompiling the default RHEL 5 kernel to include the "Opteron" config option on a quad processor opteron machine? I'm currently just using the "x86_64" version of the kernel. If you look at the ".config" options with menuconfig or xconfig, it shows that th

Re: [Beowulf] Barcelona vs. Woodcrest, computational chemistry research

2007-09-26 Thread richard . walsh
-- Original message -- From: Joe Landman <[EMAIL PROTECTED]> > Hi Andrew: > > andrew holway wrote: > > Multi reference quantum chem. Gaussian and Molcas. > > Without doing a benchmark it will be quite hard to indicate which one > may be faster. > > Barcelona is a

Re: [Beowulf] Alpha beowulf: True64 or Linux?

2007-09-26 Thread Greg Lindahl
On Wed, Sep 26, 2007 at 07:40:02AM +0100, Tim Cutts wrote: > I've heard of people making clusters out of Alphas running Linux > before - can't remember where, but I did see someone had done it. CPlant at Sandia, 2 generations of cluster at FSL in Boulder, various US spooks, etc etc. I still ha

Re: [Beowulf] Barcelona vs. Woodcrest, computational chemistry research

2007-09-26 Thread Joe Landman
Hi Andrew: andrew holway wrote: Multi reference quantum chem. Gaussian and Molcas. Without doing a benchmark it will be quite hard to indicate which one may be faster. Barcelona is a quad core with a good memory system (relative to the Clovertown). If your problems are large enough to spi

Re: [Beowulf] OT: Any beowulfers heading to Ohio Linux Fest?

2007-09-26 Thread Peter St. John
Oh I was oblivious to this. Yeah I live in Columbus, I'll stop by. Peter On 9/26/07, Joe Landman <[EMAIL PROTECTED]> wrote: > > See http://www.ohiolinux.org/ > > We will be there at a table (with a fast rabbit on it), so if you are > going, please do stop by and say hi. > > Joe > > -- > Joseph La

Re: [Beowulf] Barcelona vs. Woodcrest, computational chemistry research

2007-09-26 Thread richard . walsh
"andrew holway" <[EMAIL PROTECTED]> wrote: > > I am looking at chips for a small 20 node cluster for computational > chemistry. I'm trying to get my head around the various merits of the > different ways the cache work. > > Considering the fairly chunky bits of data that are going to be goin

[Beowulf] Re: Beowulf Digest, Vol 43, Issue 37

2007-09-26 Thread David Mathog
Tim Cutts <[EMAIL PROTECTED]> wrote > I've heard of people making clusters out of Alphas running Linux > before - can't remember where, but I did see someone had done it. I used to have a 9 node DS10 cluster running linux. Replaced it ages ago with a 20 node Athlon MP 2200+ cluster. Each rep

[Beowulf] OT: Any beowulfers heading to Ohio Linux Fest?

2007-09-26 Thread Joe Landman
See http://www.ohiolinux.org/ We will be there at a table (with a fast rabbit on it), so if you are going, please do stop by and say hi. Joe -- Joseph Landman, Ph.D Founder and CEO Scalable Informatics LLC, email: [EMAIL PROTECTED] web : http://www.scalableinformatics.com http://jackr

Re: [Beowulf] Barcelona vs. Woodcrest, computational chemistry research

2007-09-26 Thread andrew holway
Multi reference quantum chem. Gaussian and Molcas. Andy On 26/09/2007, John Leidel <[EMAIL PROTECTED]> wrote: > What chem codes are you going to be running [NWChem?] > > > On 9/26/07, andrew holway <[EMAIL PROTECTED]> wrote: > > > > Hi, > > > > I am looking at chips for a small 20 node cluster fo

Re: [Beowulf] Barcelona vs. Woodcrest, computational chemistry research

2007-09-26 Thread Vincent Diepeveen
hi Andrew, Things largely depend upon your workload and number of sockets. For integer code, intel definitely is 50% faster in IPC and most likely will clock higher too. We were shocked to see that Sjeng at specint is having a 10% worse to 5% better (depending whether you take the base or opti

Re: [Beowulf] Barcelona vs. Woodcrest, computational chemistry research

2007-09-26 Thread John Leidel
What chem codes are you going to be running [NWChem?] On 9/26/07, andrew holway <[EMAIL PROTECTED]> wrote: > > Hi, > > I am looking at chips for a small 20 node cluster for computational > chemistry. I'm trying to get my head around the various merits of the > different ways the cache work. > > Co

Re: [Beowulf] Alpha beowulf: True64 or Linux?

2007-09-26 Thread John Leidel
Of course, there is ASCI Q at LANL. It was quite a large [compaq] Alpha w/Quadrics install running Tru64. I'm not sure what HP is charging for Tru64 these days, but you could certainly get by running something like Gentoo linux quite well. BTW, what interconnect are you planning to install? On

[Beowulf] Barcelona vs. Woodcrest, computational chemistry research

2007-09-26 Thread andrew holway
Hi, I am looking at chips for a small 20 node cluster for computational chemistry. I'm trying to get my head around the various merits of the different ways the cache work. Considering the fairly chunky bits of data that are going to be going through, does anyone have any opinion on how the new 1

Re: [Beowulf] Alpha beowulf: True64 or Linux?

2007-09-26 Thread Mark Hahn
I've heard of people making clusters out of Alphas running Linux before - can't remember where, but I did see someone had done it. my organization did (and does, actually). we had several clusters of ES40's, most running linux. one (which I admined) ran tru64. we've since coalesced all the E