Thanks, I suspected of this too.
I will reinstall sage with this command
$ /usr/bin/cpufreq-selector -g performance
$ ./sage
$ /usr/bin/cpufreq-selector -g ondemand
21.12.2010 06:50, Dima Pasechnik пишет:
to install Atlas well, you should switch to "performance" CPU mode.
"On demand" might p
to install Atlas well, you should switch to "performance" CPU mode.
"On demand" might produce inconsistent timings during the tuning
stage.
http://math-atlas.sourceforge.net/atlas_install/atlas_install.html#SECTION00032000
On Dec 18, 6:07 am, "Alexey U. Gudchenko" wrote:
> Sorry for m
Sorry for my English,
I carried out some experiments, used the system reinstalation to excuse.
Configuration:
Ubuntu 10.10 with 2.6.35-23 kernel
2 GB of memory
AMD Athlon II X2 245
Processor clock frequency is on the "On demand" mode
$ cd /home/goodok/ta/sage-4.6
$ ./make
On about 0:43
> http://math-atlas.sourceforge.net/errata.html#tol
> It says that a restart is needed in this case.
> Not from scratch, but as explained
> here:http://math-atlas.sourceforge.net/errata.html#restart
>
> So I wasn't completely right - this restart is even "warmer" than the
> one I was proposing,
On Dec 19, 7:22 pm, Dima Pasechnik wrote:
> On Dec 18, 7:58 pm, Volker Braun wrote:> The tuning
> is done in the "configure" script. I would have thought that if
>
> No, I don't think so. Atlas's INSTALL.txt says:
>
> "For the impatient, here is the basic outline:
> ***
PS. Actually, re-running make is supposed to take care of that
"VARIATION EXCEEDS TOLERENCE, RERUN WITH HIGHER REPS"
error message, IMHO...
On Dec 19, 7:22 pm, Dima Pasechnik wrote:
> On Dec 18, 7:58 pm, Volker Braun wrote:> The tuning
> is done in the "configure" script. I would have thought t
On Dec 18, 7:58 pm, Volker Braun wrote:
> The tuning is done in the "configure" script. I would have thought that if
No, I don't think so. Atlas's INSTALL.txt says:
"For the impatient, here is the basic outline:
**
mkdir my_build_dir ; cd my_bu
The tuning is done in the "configure" script. I would have thought that if
that fails and you run "make" then you'd just build ATLAS with broken
timings. Do you have a reference?
I'm still working on an updated ATLAS spkg that gives up automatic tuning
after 2 tries and then proceeds to build
I have the same sort of issues with an old laptop with Pentium M CPU
(banias 1.1GHz, http://ark.intel.com/Product.aspx?id=27600)
Actually, I think the idea of re-running Altas install from scratch,
many times, as the spkg does, makes not too much sense. Googling that
issues shows that one should r
Hi
On Dec 17, 11:52 am, Karen Bindash
wrote:
> Did you really mean Sage 3.6 or was you intending to write 4.6?
Woops. Yes, 4.6.
regards,
Jan
--
To post to this group, send an email to sage-devel@googlegroups.com
To unsubscribe from this group, send an email to
sage-devel+unsubscr...@googlegr
Did you really mean Sage 3.6 or was you intending to write 4.6?
If you did mean 3.6, then the ATLAS package would have changed a lot
since Sage 3.6.
Dave
On 12/16/10, Jan Groenewald wrote:
> Hi
>
> I had ATLAS fail to build from sage-3.6 running Ubuntu 10.04.1
> on an HP Probook 4515s with and
Hi
I had ATLAS fail to build from sage-3.6 running Ubuntu 10.04.1
on an HP Probook 4515s with and AMD CPU (yesterday).
Due to time pressure I installed a prebuilt binary instead.
Sorry, no more info.
regards,
Jan
--
.~.
/V\ Jan Groenewald
/( )\www.aims.ac.za
^^-^^
--
To p
Ignore what appears to be an empty message from me. I got a message
"Your message has been sent," but I don't recall pressing the "Send"
button.
On 21 November 2010 17:17, RegB <2regburg...@earthlink.net> wrote:
> In the Virtual box it SEEMS to be necessary to give the virtual
> machine
> more tha
On 21 November 2010 17:17, RegB <2regburg...@earthlink.net> wrote:
> Update:
> Since my last update;
> The Atlas problem seemed to have been fixed with the atlas-3.8.3.p17
> spkg, however I ran into more trouble later with matplotlib-1.0.0
> and that sent me off on a course of enquiry that led me t
Update:
Since my last update;
The Atlas problem seemed to have been fixed with the atlas-3.8.3.p17
spkg, however I ran into more trouble later with matplotlib-1.0.0
and that sent me off on a course of enquiry that led me to try
building
each component separately; BLAS, NUMPY, PYTHON, ATLAS, etc.
A
So far that is working, i.e. Atlas has completed and no other packages
seem to have tripped over any side effects.
Thanks again.
On Nov 5, 7:24 pm, Volker Braun wrote:
> You can try my new atlas spkg
> athttp://trac.sagemath.org/sage_trac/ticket/10226
> if you have a minute.
>
> SAGE_ATLAS_ARCH
I get confused with the directions about running in virtual box (-:
I think it should be OK and very possible to create virtual
machines, say within virtual box within Windows Vista and install
Linux in those VMs , then go ahead and install Sage on the Linux.
i.e. (Windows Vista (Virtual Box(UBUNT
I would like to try that.
I will report back results (good or bad).
Is there any info you would like me to collect from good/bad results ?
Thanks.
On Nov 5, 7:24 pm, Volker Braun wrote:
> You can try my new atlas spkg
> athttp://trac.sagemath.org/sage_trac/ticket/10226
> if you have a minute.
>
This may be related:
I also regularly (although not deterministically) have troubles with
ATLAS on a Turion x2 notebook. I has 4Gb RAM, but I am running Ubuntu
10.4 in VirtualBox with 1.5Gb and one CPU core accessible to it. It
always takes forever to build ATLAS on it and often it crashes. Since
You can try my new atlas spkg at http://trac.sagemath.org/sage_trac/ticket/10226
if you have a minute.
SAGE_ATLAS_ARCH=HAMMER,SSE3,SSE2,SSE1
export SAGE_ATLAS_ARCH
sage -f atlas-3.8.3.p17.spkg
or just save the new spgk in SAGE_ROOT/spkg/standard/
Volker
--
To post to this group, send an email
On 11/ 5/10 05:38 PM, mhampton wrote:
Depending on what those Firefox tabs were, they could have put a
significant load on your CPU (Flash ads for example) or used enough
memory to make atlas work too hard. It would be interesting to know
if it works if you aren't running anything else.
What is
Well, I'm certainly under qualified to evaluate your thread (-:
It would SEEM that you were getting error messages that
said the flags to the compiler were wrong/inappropriate (GUESSING) ?
I got specific tune errors 2 and 255.
I'm not saying they don't have some common cause, just that the
symptoms
On 11/05/2010 11:23:01 AM, RegB wrote:
Thanks,
I have ads blocked, fairly sure that none were playing any U-tube or
other graphics,
no adobe flash.
Looking at System monitor, barely a Gig of memory is being used and
swap is NOT
being used at all. There is 4 Gig of ram.
CPU #2 is running at 100%
Thanks,
I have ads blocked, fairly sure that none were playing any U-tube or
other graphics,
no adobe flash.
Looking at System monitor, barely a Gig of memory is being used and
swap is NOT
being used at all. There is 4 Gig of ram.
CPU #2 is running at 100% most of the time during the atlas build,
Depending on what those Firefox tabs were, they could have put a
significant load on your CPU (Flash ads for example) or used enough
memory to make atlas work too hard. It would be interesting to know
if it works if you aren't running anything else.
What is your processor and amount of RAM?
-M.
25 matches
Mail list logo
