f



parallel computing on a six-core local computer

Hi, 
I just bought a six-core desktop (12 Threads) and discovered that the maximum worker allowed by the parallel computing toolbox is eight workers.  This is really disappointing, and I am just wondering if there is anyway to fully utilize the 12 processes and have 12 workers on one local machine.  I browsed the help guide for the Distributed Computing Server toolbox and it seems it only works when you'd like to create workers on remote computers. 

Your help is greatly appreciated.  Thank you!
richard
0
Richard
5/1/2010 7:18:05 PM
comp.soft-sys.matlab 211266 articles. 24 followers. lunamoonmoon (257) is leader. Post Follow

21 Replies
1365 Views

Similar Articles

[PageSpeed] 44

"Richard Liu" <richardkailiu@gmail.com> wrote in message 
news:hrhupd$2v3$1@fred.mathworks.com...
> Hi, I just bought a six-core desktop (12 Threads) and discovered that the 
> maximum worker allowed by the parallel computing toolbox is eight workers.

I believe that is the correct behavior, assuming you have just Parallel 
Computing Toolbox.  To use more than 8 local workers, or to use workers 
across multiple machines, you will need MATLAB Distributed Computing Server 
as well.

>  This is really disappointing, and I am just wondering if there is anyway 
> to fully utilize the 12 processes and have 12 workers on one local 
> machine.  I browsed the help guide for the Distributed Computing Server 
> toolbox and it seems it only works when you'd like to create workers on 
> remote computers.

That is not the case.  Could you post the URL of the documentation page that 
gave you that impression, so that our documentation staff can clarify the 
relationship between Parallel Computing Toolbox and MATLAB Distributed 
Computing Server there?

-- 
Steve Lord
slord@mathworks.com
comp.soft-sys.matlab (CSSM) FAQ: http://matlabwiki.mathworks.com/MATLAB_FAQ 


0
Steven
5/3/2010 2:10:39 PM
Steven,
Thank you so much.  Here is the URL of the documentation page I browsed: 
http://www.mathworks.com/access/helpdesk/help/pdf_doc/mdce/mdce.pdf

Maybe I am missing something here. But the discussion in the documentation seems to be concerned with server and remote workers.  The case here is that I have one local computer with the potential of running 12 workers max.   

I look forward to hearing from you. 
richard

"Steven Lord" <slord@mathworks.com> wrote in message <hrmlgt$cfj$1@fred.mathworks.com>...
> 
> "Richard Liu" <richardkailiu@gmail.com> wrote in message 
> news:hrhupd$2v3$1@fred.mathworks.com...
> > Hi, I just bought a six-core desktop (12 Threads) and discovered that the 
> > maximum worker allowed by the parallel computing toolbox is eight workers.
> 
> I believe that is the correct behavior, assuming you have just Parallel 
> Computing Toolbox.  To use more than 8 local workers, or to use workers 
> across multiple machines, you will need MATLAB Distributed Computing Server 
> as well.
> 
> >  This is really disappointing, and I am just wondering if there is anyway 
> > to fully utilize the 12 processes and have 12 workers on one local 
> > machine.  I browsed the help guide for the Distributed Computing Server 
> > toolbox and it seems it only works when you'd like to create workers on 
> > remote computers.
> 
> That is not the case.  Could you post the URL of the documentation page that 
> gave you that impression, so that our documentation staff can clarify the 
> relationship between Parallel Computing Toolbox and MATLAB Distributed 
> Computing Server there?
> 
> -- 
> Steve Lord
> slord@mathworks.com
> comp.soft-sys.matlab (CSSM) FAQ: http://matlabwiki.mathworks.com/MATLAB_FAQ 
> 
0
Richard
5/3/2010 2:27:23 PM
You probably won't get much more performance from 12 threads than 6 since there are only 6 physical cores (I'm guessing this is the latest i7 processor).

If I recall the difference is only a couple percent, best-case.
0
John
5/3/2010 2:28:06 PM
"Richard Liu" <richardkailiu@gmail.com> wrote in message 
news:hrmmgb$j5r$1@fred.mathworks.com...
> Steven,
> Thank you so much.  Here is the URL of the documentation page I browsed: 
> http://www.mathworks.com/access/helpdesk/help/pdf_doc/mdce/mdce.pdf
>
> Maybe I am missing something here. But the discussion in the documentation 
> seems to be concerned with server and remote workers.  The case here is 
> that I have one local computer with the potential of running 12 workers 
> max.
> I look forward to hearing from you. richard

Okay ... do you remember what page (or even what chapter) of that document 
gave you the impression that MATLAB Distributed Computing Server was only 
necessary for cross-machine clusters so that I can give the documentation 
staff somewhere to focus their clarification work?

-- 
Steve Lord
slord@mathworks.com
comp.soft-sys.matlab (CSSM) FAQ: http://matlabwiki.mathworks.com/MATLAB_FAQ 


0
Steven
5/3/2010 2:59:36 PM
Yes, it is the latest i7 processor.  I now have 8 workers running, and the cpu usage is 67%.  So I guess there is still room for improvement. 
"John " <John@REMOVEMEtwisthink.com> wrote in message <hrmmhl$lqb$1@fred.mathworks.com>...
> You probably won't get much more performance from 12 threads than 6 since there are only 6 physical cores (I'm guessing this is the latest i7 processor).
> 
> If I recall the difference is only a couple percent, best-case.
0
Richard
5/3/2010 3:28:06 PM
Steven, 

Chapter 2 of that document.  My impression of the whole instruction is that it's about how to run parallel programming using multiple computers.  I'd be interested in some instructions on how to do it on one local computer which could operate beyond the 8 worker limits set by the parallel computing toolbox.
Again, I could be missing something here...

"Steven Lord" <slord@mathworks.com> wrote in message <hrmocn$qg8$1@fred.mathworks.com>...
> 
> "Richard Liu" <richardkailiu@gmail.com> wrote in message 
> news:hrmmgb$j5r$1@fred.mathworks.com...
> > Steven,
> > Thank you so much.  Here is the URL of the documentation page I browsed: 
> > http://www.mathworks.com/access/helpdesk/help/pdf_doc/mdce/mdce.pdf
> >
> > Maybe I am missing something here. But the discussion in the documentation 
> > seems to be concerned with server and remote workers.  The case here is 
> > that I have one local computer with the potential of running 12 workers 
> > max.
> > I look forward to hearing from you. richard
> 
> Okay ... do you remember what page (or even what chapter) of that document 
> gave you the impression that MATLAB Distributed Computing Server was only 
> necessary for cross-machine clusters so that I can give the documentation 
> staff somewhere to focus their clarification work?
> 
> -- 
> Steve Lord
> slord@mathworks.com
> comp.soft-sys.matlab (CSSM) FAQ: http://matlabwiki.mathworks.com/MATLAB_FAQ 
> 
0
Richard
5/3/2010 3:35:20 PM
"Richard Liu" <richardkailiu@gmail.com> wrote in message 
news:hrmqfo$iq8$1@fred.mathworks.com...
> Steven,
> Chapter 2 of that document.  My impression of the whole instruction is 
> that it's about how to run parallel programming using multiple computers. 
> I'd be interested in some instructions on how to do it on one local 
> computer which could operate beyond the 8 worker limits set by the 
> parallel computing toolbox.
> Again, I could be missing something here...

I don't know where those instructions are, so I'm afraid the only suggestion 
I have is for you to contact our Technical Support staff.  They should be 
able to help you determine what you need to do to set things up.

But I will send your feedback on the documentation to the documentation 
staff for their consideration.

-- 
Steve Lord
slord@mathworks.com
comp.soft-sys.matlab (CSSM) FAQ: http://matlabwiki.mathworks.com/MATLAB_FAQ 


0
Steven
5/4/2010 2:26:57 PM
You may want to try using 6 workers since you have only 6 physical cores.

---Bob.

"Richard Liu" <richardkailiu@gmail.com> wrote in message 
news:hrmq26$k0l$1@fred.mathworks.com...
> Yes, it is the latest i7 processor.  I now have 8 workers running, and the 
> cpu usage is 67%.  So I guess there is still room for improvement. "John " 
> <John@REMOVEMEtwisthink.com> wrote in message 
> <hrmmhl$lqb$1@fred.mathworks.com>...
>> You probably won't get much more performance from 12 threads than 6 since 
>> there are only 6 physical cores (I'm guessing this is the latest i7 
>> processor).
>>
>> If I recall the difference is only a couple percent, best-case.
> 


0
Bobby
5/5/2010 1:55:54 PM
"Richard Liu" <richardkailiu@gmail.com> wrote in message <hrhupd$2v3$1@fred.mathworks.com>...
> Hi, 
> I just bought a six-core desktop (12 Threads) and discovered that the maximum worker allowed by the parallel computing toolbox is eight workers.  This is really disappointing, and I am just wondering if there is anyway to fully utilize the 12 processes and have 12 workers on one local machine.  I browsed the help guide for the Distributed Computing Server toolbox and it seems it only works when you'd like to create workers on remote computers. 
> 
> Your help is greatly appreciated.  Thank you!
> richard

I am in the same boat. This is very disappointing and really stupid of matlab to limit the amount of workers on a local computer, even if you have the parallel toolbox. Seems like a money grubbing scheme to me. 

I currently have a dual 6-core workstation (12 cores - 24 threads) and i can only use 8 threads. If i turn hyperthreading off i can use 8 cores, but thats still 4 cores im not using. I didnt pay $1500 more for the extra 4 cores to not be able to use them because of a purposeful software limitation. 

Kind of makes me want to go back to using fortran and use openMP which is free and is much more feature rich. 
0
Adnan
6/3/2010 5:48:21 PM
I just checked the pricing for a 32 worker distributed server toolbox. $21.5k. Nice mathworks. Nice. Way to screw the small business users who want to use local workstations. 

I will be using fortran from now on. 
0
Adnan
6/3/2010 6:00:25 PM
I too would like to know how to use the distributed computing toolbox for multi-processors instead of distributed machines.
0
Jon
6/4/2010 10:00:23 PM
I've actually played around with trying to discern the effects of hyperthreading in my i3 based system. It turns out that it really depends on what type of work your workers are actually doing.

HT really only seems to have a benefit if your processes include i/o operations such as disk read/writes. If you think about how HT works, you have 2 instruction pipelines for each physical core, so if one of the pipelines stalls while waiting for a slow i/o operation, it switches over to the other pipeline and keeps the core occupied.

However, if you are doing big computations that only rely on memory ops (and esp. if your dataset fits in the cache) the pipelines don't stall for an appreciable amount of time. I'm not sure what platform you are using, but on Win7 x64 (which was supposed to have an improved task scheduler over previous versions) the extra overhead that HT creates actually slows down the overall runtime in certain situations.

Basically, I'm agreeing with Bob here. You should try 8 vs. 4 workers (you should make sure to set the affinity of each worker manually as to compare apples to apples) and see if HT is actually helping or hurting.

"Bobby Cheng" <bcheng@mathworks.com> wrote in message <hrrtda$9rh$1@fred.mathworks.com>...
> You may want to try using 6 workers since you have only 6 physical cores.
> 
> ---Bob.
0
Chris
8/25/2010 8:09:04 PM
"Richard Liu" <richardkailiu@gmail.com> wrote in message <hrhupd$2v3$1@fred.mathworks.com>...
> Hi, 
> I just bought a six-core desktop (12 Threads) and discovered that the maximum worker allowed by the parallel computing toolbox is eight workers.  This is really disappointing, and I am just wondering if there is anyway to fully utilize the 12 processes and have 12 workers on one local machine.  I browsed the help guide for the Distributed Computing Server toolbox and it seems it only works when you'd like to create workers on remote computers. 
> 
> Your help is greatly appreciated.  Thank you!
> richard

Disable hyperthreading in the BIOS. Then you're down to 6 threads. This doesn't mean that you're only using half the CPU, just that the board doesn't create two 'virtual cores' per crore. I've benchmarked with and without hyperthreading on an i7 920 and without hyperthreading is faster in Matlab. Having 12 threads is not the best idea anyway if each has its own memory pool. 
0
Patrick
8/25/2010 8:23:21 PM
Hey guys, I just had a crazy idea. I don't have a 6 core intel to test this on, but if your processes are truly parallel (and hyperthreading actually improves performance) why couldn't you just run 2 instances of your main program?

I tried it on the dual core I'm currently at and have 2 instances running each with 2 workers. They don't seem to interfere with each other and as far as I can tell there is no communication between the 2 main instances for them to check that you already have open workers.

It certainly isn't elegant, but it could be a cheap workaround!

Chris
0
Chris
8/25/2010 8:28:22 PM
"Chris T" <twsic0@gmail.com> wrote in message <i53ud6$5m0$1@fred.mathworks.com>...
> Hey guys, I just had a crazy idea. I don't have a 6 core intel to test this on, but if your processes are truly parallel (and hyperthreading actually improves performance) why couldn't you just run 2 instances of your main program?
> 
> I tried it on the dual core I'm currently at and have 2 instances running each with 2 workers. They don't seem to interfere with each other and as far as I can tell there is no communication between the 2 main instances for them to check that you already have open workers.
> 
> It certainly isn't elegant, but it could be a cheap workaround!
> 
> Chris

I just got a beautiful new 12 core (2 x 2.66 Xeon Westmere), and quickly came across this issue. An effective, albiet not perfect, workaround is just as Chris indicates: open multiple instances with local 8 configurations. I have tested 1,2, and 3 instances and the sweet spot appears to be 2 for my applications. I will test additional configurations in short time.

I really hope MathWorks will increase the local configuration limit to accommodate the increasing number of cores available on the consumer (albiet high-end) market.

Cheers,
Dave
0
David
9/27/2010 4:32:07 PM
I have a dual 6-core (Xeon) PC.

How do I run two instances of Matlab, each restricted on a single processor?

I would be grateful if you could indicate me how to do this both under windows and under Linux - possibly from the command line.. :-)

According to Chris' and Dave's suggestions, this indeed seems the only way around the 8-core restriction, imposed by Mathworks, for embarassingly-parallel applications of course.

> I really hope MathWorks will increase the local configuration limit to accommodate the increasing number of cores available on the consumer (albiet high-end) market.

I fully share the same frustration and disappointment.

Thanks in advance for your help.
Yours
Michele
0
Michele
12/14/2010 9:57:05 AM
By the Linux command 'taskset' it might be possible as follows:

taskset -c 0-5 /usr/local/MATLAB/R2010b/bin/matlab -nodesktop -nosplash
taskset -c 6-11 /usr/local/MATLAB/R2010b/bin/matlab -nodesktop -nosplash

(launched from two separate terminal windows or ssh sessions, etc.)

Then on each of the two concurrent matlab sessions, simply invoke:

matlabpool(6);

Note: the second time I launched 'matlabpool(6)' I got the following warning:
<<<Warning: Found 1 pre-existing parallel job(s) created by matlabpool that are running.
You can use   'matlabpool close force local'    or create a configuration for the distcomp.localscheduler object and use   'matlabpool close force <configurationName>'    to remove all jobs created by matlabpool.>>>


Nevertheless from a quick-and-dirty test, I think I could get all 12 cores working independently, at least from monitoring each core by the command 'gnome-system-monitor &'.
In addition, the duration of each elementary task (i.e. a system call to a C-code calling the sleep function) suggested individual tasks went in parallel.

****************************************************
******* I would appreciate if some could confirm this ******* 
****************************************************

Thanks,
Michele
0
Michele
12/14/2010 11:58:04 AM
I have to say, I'm having a hard time understanding the frustration with the 8-worker limit, but that might be because my experience is so vastly different from the other posters.

I too have  Dual 6-core Xeon PC. With hyperthreading, I get at most a factor of 2 speed-up when running with 4-8 labs, the optimum number being about 4 labs.
Without hyperthreading, I've observed run times that (with 8 labs) are even a factor of 8 slower! 

Therefore, I have managed to convince myself that the Parallel Computing Toolbox doesn't really obtain even close to linear speed-up on local workers, and eventually it becomes counter-productive to increase the number of local labs beyond a certain optimal point. In light of that, we should even be glad that the 8-lab limit is imposed, because if we were allowed to increase the number of labs arbitrarily, we would hurt ourselves!

That seems to be consistent with the online PCT literature as well. The linear speed-up results advertised there all seem to be for the case of remote workers, perhaps because (quoting p. 2-29 from the manual) 

"on a remote cluster... workers can simultaneously create their arrays, saving transfer time"

 Anyway, I would be interested to hear it confirmed that people have actually managed to get close to an 8-fold speed-up on local workers, my experience being as different as it is.
0
Matt
12/14/2010 3:47:22 PM
Thank you all for commiserating on the ridiculous local scheduler cap. I think Mathworks is a bit behind the capabilities of HW workstations. The speedup and benefits of parallelization are very context-specific and have a lot to do with the way you have decomposed your problem, but in theory you should be able to prototype a highly parallel algorithm on a local scheduler and scale it up to the cluster "seemlessly," as Mathworks denotes in its documentation.

That said, here is something I have found to get around the local scheduler limitation in a way. There is no limit on how many local schedulers you can open. HOWEVER, you need to trick Matlab into thinking that you actually have different local schedulers. If you open a local scheduler in two different sessions, you'll find that you're tied to the same scheduler with the same data location. Change the .DataLocation property to some new directory (it has to be one that exists, or Matlab barfs at you), and you have a new scheduler. I've tried this on a shared storage cluster where I've run two simultaneous Matlab sessions, each with 8 cores dedicated to them. Under default conditions, no dice... each session is tied to the same 8-core scheduler. If I change the .DataStorage on one of the schedulers, voila, I can run another 8-task job simultaneously on the second scheduler. 

"Michele Giugliano" wrote in message <ie7m4c$kue$1@fred.mathworks.com>...
> By the Linux command 'taskset' it might be possible as follows:
> 
> taskset -c 0-5 /usr/local/MATLAB/R2010b/bin/matlab -nodesktop -nosplash
> taskset -c 6-11 /usr/local/MATLAB/R2010b/bin/matlab -nodesktop -nosplash
> 
> (launched from two separate terminal windows or ssh sessions, etc.)
> 
> Then on each of the two concurrent matlab sessions, simply invoke:
> 
> matlabpool(6);
> 
> Note: the second time I launched 'matlabpool(6)' I got the following warning:
> <<<Warning: Found 1 pre-existing parallel job(s) created by matlabpool that are running.
> You can use   'matlabpool close force local'    or create a configuration for the distcomp.localscheduler object and use   'matlabpool close force <configurationName>'    to remove all jobs created by matlabpool.>>>
> 
> 
> Nevertheless from a quick-and-dirty test, I think I could get all 12 cores working independently, at least from monitoring each core by the command 'gnome-system-monitor &'.
> In addition, the duration of each elementary task (i.e. a system call to a C-code calling the sleep function) suggested individual tasks went in parallel.
> 
> ****************************************************
> ******* I would appreciate if some could confirm this ******* 
> ****************************************************
> 
> Thanks,
> Michele
0
mayosten (2)
4/26/2011 5:07:04 AM
"Peter" wrote in message <ip5jto$ml7$1@fred.mathworks.com>...
> Thank you all for commiserating on the ridiculous local scheduler cap. I think Mathworks is a bit behind the capabilities of HW workstations. The speedup and benefits of parallelization are very context-specific and have a lot to do with the way you have decomposed your problem, but in theory you should be able to prototype a highly parallel algorithm on a local scheduler and scale it up to the cluster "seemlessly," as Mathworks denotes in its documentation.
> 
> That said, here is something I have found to get around the local scheduler limitation in a way. There is no limit on how many local schedulers you can open. HOWEVER, you need to trick Matlab into thinking that you actually have different local schedulers. If you open a local scheduler in two different sessions, you'll find that you're tied to the same scheduler with the same data location. Change the .DataLocation property to some new directory (it has to be one that exists, or Matlab barfs at you), and you have a new scheduler. I've tried this on a shared storage cluster where I've run two simultaneous Matlab sessions, each with 8 cores dedicated to them. Under default conditions, no dice... each session is tied to the same 8-core scheduler. If I change the .DataStorage on one of the schedulers, voila, I can run another 8-task job simultaneously on the second scheduler. 
======================

Ah, and now that you've advertised this, how long do you suppose it will be before TMW blocks this in future releases :-)

Anyway, I'm curious, is there any way that non-advanced parallel programmers can take advantage of this loophole? A certain portion of the PCT is geared toward people like me, who have no idea what a scheduler is and how they're used. For example, is there a way to extend a simple PARFOR loop to run as if on a pool of >8 labs?

parfor i=1:K
 ...
end
0
mattjacREMOVE (3196)
4/26/2011 9:08:04 AM
"Matt J" wrote in message <ip621k$782$1@fred.mathworks.com>...
> ======================
> 
> Ah, and now that you've advertised this, how long do you suppose it will be before TMW blocks this in future releases :-)
> 
> Anyway, I'm curious, is there any way that non-advanced parallel programmers can take advantage of this loophole? A certain portion of the PCT is geared toward people like me, who have no idea what a scheduler is and how they're used. For example, is there a way to extend a simple PARFOR loop to run as if on a pool of >8 labs?
> 
> parfor i=1:K
>  ...
> end

Haha, good point!

As for creating pools of > 8 labs, I'm unaware of any way to magically lift the 8 worker limit. By default, Matlab is going to use a local scheduler, and the local scheduler is where that 8-worker limit comes into play. If you're at an institution with a license for Matlab Distributed Computing Server, you could specify a third-party scheduler that supports more simultaneous workers, but there again, we're getting further down into the weeds on PCT. Without doing something illegal, I'm not sure there's any way to just lift the 8-worker limit for things like matlabpool and parfor.
0
mayosten (2)
4/26/2011 10:25:04 PM
Reply:

Similar Artilces:

parallel computing, distributed computing
HI, I have an optimization code which has some part code is common and required as inputs for the two optimization routines within. However once this is done, the remaining could be run parallely independednt of each other and imrovized on the speed and finally the two results could be fetched in one single main optimization file and compared. Though i feel this could be done, i do not knopw the exact procedure as well do not know if there is some setting to be done in matlab and modify my main code as well. Please note i have matlab 2007b and distributed computing toolbox v3.2. I...

Call For Papers: Soft Computing for Bioinformatics (SOFT COMPUTING Journal
=20 SOFT COMPUTING Journal - SPRINGER =20 Special Issue on =20 Soft Computing for Bioinformatics =20 CALL FOR PAPERS =20 The past few years have witnessed phenomenal growth of=20 bioinformatics, an exciting field devoted to the interpretation and=20 analysis of biological data using computational techniques. Among=20 the large number of computational techniques used, soft computing,=20 which incorporates=20 =20 * neural networks,=20 * evolutionary computation,=20 * fuzzy systems, and=20 * chaos,=20 =20 stands out because of its demonstrated strength in handling=20 imprecise information and providing novel solutions to hard=20 problems.=20 =20 This special issue aims at not only showcasing innovative=20 applications of soft computing techniques to bioinformatics, but=20 also clarifying outstanding issues for future progress.=20 =20 Biological areas of interest include but are not limited to the=20 following:=20 =20 * protein structure and function,=20 * genomics,=20 * proteomics,=20 * molecular sequence analysis,=20 * evolution and phylogenetics,=20 * molecular interactions and structure,=20 * gene expression,=20 * metabolic pathways,=20 * regulatory networks,=20 * developmental control and systems biology. =20 =20 Papers should be submitted in PDF format via email to any of the=20 following guest editors by 30 March 2004:...

parallel computing in distributed computing server
Hi, there, Is there some way that can use the parallel computing code such as parfor or spmd in the distributed computing server? I currently ran a test for comparing the difference between the parallel toolbox with in stand alone mode in a computer with two CPUs, and the distributed computing server in two computers(each of them using just one CPU). I found the computing capabilities are pretty much the same with each other. I was curious how to run a parallel computing in the distributed server. Thanks, Wei Wei, The Parallel Computing Toolbox (PCT) serves two purposes: leverage...

Multicore distributed computing
I'm running R2008B on a 2 PC cluster (XP Pro). Each PC has four cores. I've confirmed that the Matlab distributed computing toolboox and parallel computing toolbox are properly installed and running on both machines. 4 green checks appear when I validate through the configuration manager. Max number of workers set to 16. My question: I want to confirm that when I run a job all 4 cores on each machine are utilized for a total of 8 cores. My error when I try to open a pool with 8 cores is below. Is there another command I should try? Thanks. >> matlabpool open 8 Starti...

matlab for parallel computing
Is any body doing parallel computing with matlab here? ...

parallel computing with matlab
Hello friends, I want to know that ,is it possible to do parallel computing with MATLAB on a single computer.I want that different modules communicate and interlinked with each other so that called by other functions also. Hi, have a look at the distributed computing toolbox: http://www.mathworks.com/products/distribtb/ Although mainly intended to seperate the workload to different machines you may also run it on one machine... Titus www.mathworks.com\products\dis "shivi chaturvedi" <shivi79@rediffmail.com> schrieb im Newsbeitrag news:ef3d35a.-1@webcrossing...

Local computer name dissappears from workgroup computers
After a day of trying to connect to various hotspots, some successfully and some unsuccessfully, my laptop (IBM ThinkPad R50) running XP Pro SP1 fails to display its own name in Workgroup Computers. The larger problem is that it no longer wants to share files in my home local area network It will connect to my access point for purpose of internet access fine. The solution has been to take the laptop and connect it to a large network with many different workgroups and servers. That solution is now practical. However, this scenario is repeatable. I do not know what actually triggers the problem. It does seem to be related to unsuccessful connections. Anyone have any thoughts? ...

parallel computing toolbox without distributive computing toolbox
Hi, Is it true that the parallel computing toolbox won't work if the distributive computing toolbox has not been installed? Because then if I type in 'matlabpool', I received the following err msg: " License checkout failed. License Manager Error -4 Maximum number of users for Distrib_Computing_Toolbox reached. Try again later. To see a list of current users use the lmstat utility or contact your License Administrator. Troubleshoot this issue by visiting: http://www.mathworks.com/support/lme/R2008b/4 Diagnostic Information: Feature: Distrib_Computing_Toolbox License pa...

Computer to Computer chat
Hello, I have just started understanding WinSock and I'm wondering how to make a program talk to another computer through the internet? I have the other's IP address, though both computers are on networks (I couldn't seem to get any example codes to work with my computer, I think it was because of the router...) I know the port that I will be talking through and listening on. I just want this program to send simple messages that will be understood on the other side. i.e let p! = program 1 let P2 = program 2 P1 -> P2: msg = TALK P2 -> P1: msg = CONFIRM P1 -> P2...

Parallel Computing
Hi all, I'm thinking of getting a workstation with two 6-core processors (i.e., 12 physical cores in total, each of which can take 2 threads, so 24 of them will appear on Windows Task Manager), but I don't know whether the MATLAB's Parallel Computing Toolbox (PCT) supports as many. As far as I know, the configuration for PCT only allows for maximum of 8 workers (Cluster size). Any suggestion for me to make use of more resource available? Regards, Jason "Jason Park" <jason.park@buseco.monash.edu.au> writes: > I'm thinking of getting a workstation...

Computer to Computer chat
Hello, I have just started understanding WinSock and I'm wondering how to make a program talk to another computer through the internet? I have the other's IP address, though both computers are on networks (I couldn't seem to get any example codes to work with my computer, I think it was because of the router...) I know the port that I will be talking through and listening on. I just want this program to send simple messages that will be understood on the other side. i.e let p! = program 1 let P2 = program 2 P1 -> P2: msg = TALK P2 -> P1: msg = CONFIRM P1 -> P2...

Parallel programming on multi-core computers?
I would like to explore parallel programming on the Intel multi-core architecture (Core 2 Due, quad-core Q6600, etc.) on Linux and maybe Windows. I have a lot of parallel programming experience from my grad school days in MPI, Java threads, and Pthreads, but I was wondering what the state-of-the-art parallel programming library for these Intel architectures? I've heard about Intel Threading Building Blocks (TBB) but only in passing. ...

Matlab Student Version and Parallel computing toolbox
I am planning to order the Matlab student version and the parallel computing toolbox. I have access to an Nvidia cuda enabled GPU card.Does the parallel computing toolbox (student version) support GPU computing? I thought I will ask before placing my order. Many Thanks for any help. H On 01/01/11 11:22 PM, Hamdi wrote: > > I am planning to order the Matlab student version and the parallel > computing toolbox. > > I have access to an Nvidia cuda enabled GPU card.Does the parallel > computing toolbox (student version) support GPU computing? > > I tho...

computer 2008 model for soft ware sys..
wf fefh dfkl http://www.freewebs.com/thuiss/ http://indianfriendfinder.com/go/g906725-pmem ...

Recent soft computing papers and MATLAB Toolboxes
Dear Colleagues, I would like to call your kind attention to the updated website of the Soft Computing Research Group at the University of Veszprem (Hungary) http://www.fmt.vein.hu/softcomp/ You can download MATLAB Toolboxes: - Fuzzy Clustering MATLAB Toolbox - Genetic Programming MATLAB Toolbox - Interactive Evolutionary Strategy (EASy) MATLAB Toolbox - Constrained Fuzzy Model Identification for the FMID Toolbox independent MATLAB programs related to: - Data mining * Fuzzy clustering based time-series segmentation * Supervised Fuzzy Clustering for the Identification of Fuzzy Classifiers * Fuzzy Modeling with Multidimensional Membership Functions: Grey-Box Identification and Control Design * Compact TS-Fuzzy Models through Clustering and OLS plus FIS Model Reduction * Inconsistency Analysis of Labeled Data * Star plots - MATLAB files for Graphical Representation of trace elements of clinkers - Process control and monitoring * Feedback Linearizing Control Using Hybrid Neural Networks Identified by Sensitivity Approach * Incorporating Prior Knowledge in Cubic Spline Approximation - Application to the Identification of Reaction Kinetic Models * Identification and Control of Nonlinear Systems Using Fuzzy Hammerstein Models - A Simple Fuzzy Classifier based on manuscripts in PDF about - fuzzy model based process control and monitoring - fuzzy clustering and classification - incorpor...

Does Matlab provide function to compute local histogram equalization?
Hello: Does Matlab provide functiont to compute local histogram equalization? I know it provides histeq "Enhance contrast using histogram equalization" Thank you -Daniel What do mean by local? a square of a pictrue, would not be hard to implement? /Anders "Daniel Mark" <d@hotmail.com> skrev i meddelandet news:ef19822.-1@webx.raydaftYaTP... > Hello: > > Does Matlab provide functiont to compute local histogram > equalization? > > I know it provides histeq "Enhance contrast using histogram > equalization" > > Thank you > -Dan...

Matlab 2011a student version/ Parallel computing box
Hi, Does the parallel tool box version 4.3 support GPU computing? If not, how to get GPU computing capanility with Matlab 2011a student version? I thought I will order the Matlab 2011a student but the parallel tool box that is available to students (v 4.3) does not mention GPU computing. I will appreciate any advice. Thank you, H You might consider the student version of Jacket, here: http://www.accelereyes.com See comparison description: http://www.accelereyes.com/products/compare "Hamdi " <hamdi.mani@asu.edu> wrote in message <inoni0$fh7$1@fred.mathworks.com>....

Using Internet tool kit how to communicate with 2 computers which are far apart and each computer is part of local network
Hi, &nbsp; I started using internet tool kit very recently. I have an application as briefed below, Computer 1 is connected to local network 1. Computer 2 is connected to&nbsp;Remote network 2. Network 1 and Remote network 2 are far apart and they are connected to internet. It is required to open a Telnet session between computer 1 and 2 through internet. How to achieve this? Which are the functions to be used? Where to enter Gateway&nbsp; and Subnet mask&nbsp;addresses? Can any body help me out. &nbsp; Regards &nbsp; SajK &nbsp; &nbsp; ...

Computational Mathematics and Computer Science
Hi. I'm trying to decide between colleges and one of the schools on my list doesn't offer a major in computer science (although a minor is offered) but offers a major in 'Computational Mathematics'. Is COMPUTational Mathematics related anyhow with the more theoretical COMPUTer Science ? Is there any connection between these fields of study (besides the first six letters) ? Also, if I only have a minor in Computer Science is it still possible for me to apply to graduate school (Phd) for Computer Science ? How do admission officers at Universities see such people who major in on...

parallel computing
Hi All, While working on a project, I discovered lots of little opportunities for real parallelism. For instance, the following class initialization: from pg import DB class example: def __init__(self): # find somehow HOST1, HOST2 self.member1=DB('database1',host=HOST1).query("SELECT...").getresult() self.member2=self.my_aux_func() # some more processing here self.member3=DB('database1',host=HOST2).query("SELECT...").getresult() # other things here will ask other physical computers to do some of the work... and wait for the resu...

Use another Computer for computation
Hi, I've two computer: -an old portable (A); -a new gen desktop (B); Is there a way to write the code in the computer A and the do the computation with computer B trought internet or wi-fi? (Think computer B as a server and A as a client). Thanks. ...

Computer Service, Custom Computers
I am a system builder, certified computer technician that can build a computer to your specifications, Office computers, gaming computers even network servers! I offer door-to-door service for computer servicing, upgrades and repairs. Network installation whether it be wired or wireless. No job is too little or too big. Please visit my website at http://www.cpu-guy.com or call (845) 246-3348. I am located in the Saugerties, New York area and provide computer help throughout the hudson valley and beyond. Computer components are also available at great prices. My service rates ar...

Airport computer to computer problems
I have a new laptop, and am trying once again to get the computer to computer network going. This has been a vary painfull process in the past two times in the past. I accidently turned off airport and can not get it set up once again. Somehow when you turn airport off it resets all the parameters. When I go to create a network I get the error "There was an error creating your computer to computer network". Thats it, no error number, nothing. Did Bill Gates design this? Roger << When I go to create a network I get the error "There was an error creating your computer to...

Parallel computations
Hello, I have no experience with Core 2 Duo or Core 2 Quad processors. It possible to use them to parallel computations, say using Personal Grid? Thanks for any advice. Stefan Porubsky Hi, yes Regards Jens Stefan Porubsky wrote: > Hello, > > I have no experience with Core 2 Duo or Core 2 Quad processors. It > possible to use them to parallel computations, say using Personal Grid? > > Thanks for any advice. > > Stefan Porubsky > > if you are a fortran or C programmer, a multi-core processor will benefit you. so does ...

Web resources about - parallel computing on a six-core local computer - comp.soft-sys.matlab

Computing - Wikipedia, the free encyclopedia
For the formal concept of computation, see computation . For the magazine, see Computing (magazine) . For the scientific journal, see Computing ...

Network Computing
Through a For IT, By IT editorial filter, Network Computing connects the dots between architectural approach and how technology impacts the business, ...

Apple Signs Deal To Use Google’s Cloud Computing
More and more businesses are moving their computing platform to the cloud. Cloud computing is the term given to accessing computer and database ...

The coolest thing Sony has built in years might show us the future of computing
... the coolest examples of this back in October when we told you about the MUV Bird, which you can read about here . While the idea that our computing ...

Seattle’s Tableau Software snaps up database-computing startup in Germany
Seattle’s Tableau Software has acquired HyPer, a database-computing startup that spun out of research at a university in Munich, Germany. As ...

Here's how much computing power Google DeepMind needed to beat Lee Sedol at Go
Google DeepMind may have made history by beating the world champion of Chinese board game Go on Wednesday but it needed an awful lot of computing ...

Skyport Raises $30M from GV and Cisco Investments to Ramp Delivery of Zero-Trust Secure Computing Infrastructure ...
Skyport Systems, the company redefining enterprise security architecture, today announced a $30 million round of funding, enabling the company ...

Google throws its computing brains into tackling the Zika virus
The tech giant hopes to help Unicef and other nonprofits ambush the virus by predicting where to focus resources. Oh, and it's throwing in a ...

2015's 'Nobel Prize of computing' honors encryption pioneers
Unless you've just finished an incredibly dedicated Rip Van Winkle cosplay session, you're probably well aware of how hot a topic encryption ...

Hybrid Cloud Spurs IoT, Cognitive Computing, More: IBM
A new IBM report shows leading enterprises using hybrid cloud as a launching point for next-generation initiatives like IoT and cognitive computing. ...

Resources last updated: 3/21/2016 7:56:51 AM