NONMEM Users Network Archive

Hosted by Cognigen

RE: CycleCloud BigScience Challenge giving away ~8-hours on 30000 core cluster for research

From: Mark Sale - Next Level Solutions <mark>
Date: Mon, 14 Nov 2011 10:14:41 -0700
Leonid,
 = ;  Very similar to my experience, running large jobs (24+ hours on one= core) on a cluster of 9, 6 core machines - typically about 80% efficiency = (by exactly the same metric).  I think that is the spectrum, ~90% on s= ingle machine, ~80% on cluster, and it looks like maybe 50-60% on cloud.&nb= sp; Only hardware configuration missing is VPN - I'd guess it would be betw= een cluster and cloud. (and might be a very good business model for small u= sers to share resources).

Your mileage may va= ry.




<= br>
Mark Sale MD
President, Next Level Solutions, LLC
www.NextLevelSolns.com
919-846= -9185
A carbon-neutral company
See our real time solar = energy production at:


-------- Original Message --------
Subject:= Re: [NMusers] CycleCloud BigScience Challenge giving away
~8-hours o= n 30000 core cluster for research
From: Leonid Gibiansky < To: Bill Knebel < on the same computer improved the speed proport= ionally to the number of
processors with the efficiency (for 12 proc= essor run) in the range of 85
to 95% for all methods except BAYES, w= hich had parallelization
efficiency of about 70%." Efficiency was de= fined as

100%*(single CPU run time)/(multiple CPU run time)/(n= umber of processors)

Roughly, the model run on 11 processors w= as 10 times faster than the
single-processor run of the same model
Leonid


------------------------------------= --
Leonid Gibiansky, Ph.D.
President, QuantPharm LLC
web:=
www.quantpharm.com
e-ma= il: LGibiansky at quantpharm.com
= tel: (301) 767 5566



On 11/14/2011 9:54 AM,= Bill Knebel wrote:
> David,
>
> Some limited be= nchmarking results are listed at the bottom of this
> email. It is= also important to remember that there are ways that
> cloud-compu= ting helps beyond parallel NONMEM. Cloud-computing allows on
> dem= and, user specific, clusters that can grow and shrink depending on
&g= t; user requirements. This allows for rapid completion of large bootstrap > processes (500 - 1000 jobs) in the time it takes to run one,
= > non-parallel job. Users can also evaluate model variants simultaneo= usly
> with cluster size being increased or decreased as needed. U= sers do not
> have to worry about competing for resources (compute= cores) with other
> users because cloud-computing clusters are us= er and/or project specific.
> The performance gains are evident in= parallel NONMEM jobs and single
> modeling projects, but it is im= portant to look beyond the simple
> benefits of individual job/pro= ject speed-up and more towards the impact
> of cloud-computing on = the entire portfolio of modeling and simulation
> projects in a gi= ven group or company.
>
> Bill
>
>
= > Model 1 - ADVAN6, 1000 subjects, dual linear and non-linear elimina= tion
> CoresRuntime (hr)
> 180.7
> 816.8
&= gt; 168.9
> 246.2
> 483.8
> 962.5
>
= > Model 2 - ADVAN6 - 70 subjects, PKPD model
> CoresRuntime = (hr)
> 14 hr
> 81.1
> 160.66
> 240.4 > 480.31
>
> cores = number of compute cores (va= lue of "NODES=" argument in NONMEM
> pmn file)
>
= > On Nov 9, 2011, at 9:28 PM, David Foster wrote:
>
>&= gt; I agree with Julia,
>>
>> Thanks for this Bill,= but quantitative benchmarking would be very much
>> appreciate= d.
>>
>> Regards,
>>
>> Dav= id
>>
>>
>> On 10/11/11 9:55 AM, "Ivash= ina, Julia" < >> Bill,
>>
>> This is nice= to hear about the speed improvements for NONMEM you
>> rec= eived.
>>
>> Could you please describe gains in= performance in a qualitative
>> manner with model examples= you used.
>> I think everyone will benefit from such NONME= M 7.2 benchmarking.
>>
>> I posted a similar qu= estion in March but not many responded.
>>
>> T= hanks,
>> Julia
>>
>> -----Orig= inal Message-----
>> From:
owner-nmusers sg: <x-msg:idssoftware oftware >
>> >
>> > William J. Bachman= , Ph.D.
>> > Director, Pharmacometrics R&D
>= ;> > Icon Development Solutions
>> > 6031 Univ= ersity Blvd., Suite 300
>> > Ellicott City, MD 21043
= >> > Office 215-616-8699
>> >
><x-msg:http://bit.ly/BigScience
>> >
= >> > The application process is simply answering 4 question= s (takes
>> less than half an hour): State who you are, wha= t is your research,
>> why it is important, and how you cur= rently run computation. The
>> applicatiopn is available he= re:
>> > http://cyclecomputing.com/big-science-challenge/overv= iew
>> >
>> > So far, response ha= s been great, and Inside HPC covered
>> descriptions of som= e of the recent applications we've received:
>> > http://insidehpc.com/2011/10/27= /24209/
>> > http://blog.cyclecomputing.com/
>> >
>&= gt; > Submissions are due by November 7th, so submit early and we ho= pe
>> to help some of you get some BigScience done quickly.=
>> >
>> > Best,
>> &= gt; Jason
>> >
>> > --
>>= >
>> >
>> > ====== ========================== ====
>> > Jason A. Stowe
>> >= cell: 607.227.9686
>> > main: 888.292.5320
>&g= t; >
>> > http://twitter.com/jasonastowe/
>> > http://twitter.com/cyclecomputing/ >> >
>> > Cycle Computing, LLC
&= gt;> > Leader in Open Compute Solutions for Clouds, Servers, and = Desktops
>> > Enterprise Condor Support and Management T= ools
>> >
>> > http://www.cyclecomputing.com <http://www.cyclecomputing.com/>;
&= gt;> > http://www.cycleclou= d.com <http://www.cycleclo= ud.com/>;
>> >
>> >
>&= gt; > --
>> > Nick Holford, Professor Clinical Ph= armacology
>> > Dept Pharmacology& Clinical Pharmaco= logy
>> > University of Auckland,85 Park Rd,Private Bag = 92019,Auckland,New
>> Zealand
>> > tel:+= 64(9)923-6730 fax:+64(9)373-7090 mobile:+64(21)46 23 53
>> = > email: //692/n.holford ; http:= //www.fmhs.auckland.ac.nz/sms/pharmacology/holford
>> &= gt;
>> >
>> >
>> >=
>> > --
>> >
>> >=
>> > ================= ==================
>> &= gt; Jason A. Stowe
>> > cell: 607.227.9686
>>= ; > main: 888.292.5320
>> >
>> &g= t; http://twitter.com/jasonastow= e/
>> > http://twitter.com/cyclecomputing/
>> >
>= > > Cycle Computing, LLC
>> > Leader in Open C= ompute Solutions for Clouds, Servers, and Desktops
>> > = Enterprise Condor Support and Management Tools
>> >
= >> > http://www.c= yclecomputing.com <htt= p://www.cyclecomputing.com/>;
>> > http://www.cyclecloud.com <http://www.cyclecloud.com/>;
>> = >
>> > ICON plc made the following annotations. >> > -----------------------------------------------------= -------------------------
>> > This e-mail transmission = may contain confidential or legally
>> privileged informati= on
>> > that is intended only for the individual or enti= ty named in the
>> e-mail address. If you
>> = > are not the intended recipient, you are hereby notified that any
= >> disclosure, copying,
>> > distribution, = or reliance upon the contents of this e-mail is
>> strictly= prohibited. If
>> > you have received this e-mail trans= mission in error, please reply
>> to the sender, so that >> > ICON plc can arrange for proper delivery, and then pl= ease delete
>> the message.
>> > Thank Y= ou,
>> > ICON plc
>> > South County B= usiness Park
>> > Leopardstown
>> > D= ublin 18
>> > Ireland
>> > Registered= number: 145835
>> >
>> >
>&= gt; >
>>
>> Notice: This e-mail message,= together with any attachments, contains
>> information of = Merck & Co., Inc. (One Merck Drive, Whitehouse Station,
>> = New Jersey, USA 08889), and/or its affiliates Direct contact
>= > information
>> for affiliates is available at
= >> http= ://www.merck.com/contact/contacts.html) that may be confidential,
= >> proprietary copyrighted and/or legally privileged. It is in= tended
>> solely
>> for the use of the indi= vidual or entity named on this message. If
>> you are
= >> not the intended recipient, and have received this message = in error,
>> please notify us immediately by reply e-mail a= nd then delete it from
>> your system.
>>
= >>
>>
>> *
>>
>> >> David Foster, PhD
>> *Lecturer
>>
= >> School of Pharmacy and Medical Sciences
>> Playford= Building
>> Room P4-08
>> City East Campus
&= gt;> University of South Australia
>> Adelaide SA 5000
= >> CRICOS Provider Number: 00121B
>> Phone: +61 8 8302= 2055
>> Fax: +61 8 8302 2389
>> Email: >
>> _
>> Note: This message (including an= y attachments) is confidential and is
>> intended only for the = individual(s) named. If you have received it by
>> mistake, ple= ase let the sender know by e-mail reply and delete it from
>> y= our system.
>> PPlease consider the environment before printing= this email
>>
>

left.letterhead
(image/png attachment: left.letterhead)

Received on Mon Nov 14 2011 - 12:14:41 EST

The NONMEM Users Network is maintained by ICON plc. Requests to subscribe to the network should be sent to: nmusers-request@iconplc.com.

Once subscribed, you may contribute to the discussion by emailing: nmusers@globomaxnm.com.