QuantLib › quantlib-users

Compilation Speed Question

Classic

List

Threaded

3 messages Options

John Orford

Compilation Speed Question

Hey Guys,

I would like to get acquainted again with compiling QuantLib and adding some Swig Python functionality (picking up where I left off a few months ago with lots of trial and error - be prepared for questions...).

However I am away from my desktop for the next month and due to home networking issues thought I might try my hand at using a Google Compute VM.

Any recommendations for the best setup?

https://cloud.google.com/products/compute-engine/#pricing

Essentially, should I expect memory or cpu bottlenecks?

Thanks!

John

------------------------------------------------------------------------------
HPCC Systems Open Source Big Data Platform from LexisNexis Risk Solutions
Find What Matters Most in Your Big Data with HPCC Systems
Open Source. Fast. Scalable. Simple. Ideal for Dirty Data.
Leverages Graph Analysis for Fast Processing & Easy Data Exploration
http://www.hpccsystems.com
_______________________________________________
QuantLib-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/quantlib-users

Luigi Ballabio

Re: Compilation Speed Question

Just one or two data points. I just compiled the wrappers on my
laptop, and it took about 7 minutes using 100% of one 2.40Ghz Intel
P8600 core. It used about 80% of 4Gb of memory.
However, I used to compile it on my previous laptop. I think it had
half as much memory, and it managed (possibly with some swapping). It
took longer that 7 minutes, but that was also because the processor
was way slower (I can't remember how much, though).
Of the machines you list, an n1-standard-1 should be enough, and a
g1-small might. You won't need more than one core anyway (the wrappres
are one big file).

Hope this helps,
Luigi

On Mon, Jun 9, 2014 at 8:00 AM, John Orford <[hidden email]> wrote:

> Hey Guys,
>
> I would like to get acquainted again with compiling QuantLib and adding some
> Swig Python functionality (picking up where I left off a few months ago with
> lots of trial and error - be prepared for questions...).
>
> However I am away from my desktop for the next month and due to home
> networking issues thought I might try my hand at using a Google Compute VM.
>
> Any recommendations for the best setup?
>
> https://cloud.google.com/products/compute-engine/#pricing
>
> Essentially, should I expect memory or cpu bottlenecks?
>
> Thanks!
>
> John
>
> ------------------------------------------------------------------------------
> HPCC Systems Open Source Big Data Platform from LexisNexis Risk Solutions
> Find What Matters Most in Your Big Data with HPCC Systems
> Open Source. Fast. Scalable. Simple. Ideal for Dirty Data.
> Leverages Graph Analysis for Fast Processing & Easy Data Exploration
> http://www.hpccsystems.com
> _______________________________________________
> QuantLib-users mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/quantlib-users
>

--
<https://implementingquantlib.blogspot.com>
<https://twitter.com/lballabio>

------------------------------------------------------------------------------
HPCC Systems Open Source Big Data Platform from LexisNexis Risk Solutions
Find What Matters Most in Your Big Data with HPCC Systems
Open Source. Fast. Scalable. Simple. Ideal for Dirty Data.
Leverages Graph Analysis for Fast Processing & Easy Data Exploration
http://www.hpccsystems.com
_______________________________________________
QuantLib-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/quantlib-users

John Orford

Re: Compilation Speed Question

Out of curiosity, this afternoon, I took an hour to fire up a 16 core 15Gb VM. It took just over ten minutes to compile the libquantlib debian source package -- excluding the tests. There's quite a few times when all cores are fully loaded.

As for the swig wrappers - you're right, mostly single core. It hits 'surfaces' and takes quite a few minutes to compile...

Thanks for the input.

On 9 June 2014 18:01, Luigi Ballabio <[hidden email]> wrote:

Just one or two data points. I just compiled the wrappers on my
laptop, and it took about 7 minutes using 100% of one 2.40Ghz Intel
P8600 core. It used about 80% of 4Gb of memory.
However, I used to compile it on my previous laptop. I think it had
half as much memory, and it managed (possibly with some swapping). It
took longer that 7 minutes, but that was also because the processor
was way slower (I can't remember how much, though).
Of the machines you list, an n1-standard-1 should be enough, and a
g1-small might. You won't need more than one core anyway (the wrappres
are one big file).

Hope this helps,
Luigi

On Mon, Jun 9, 2014 at 8:00 AM, John Orford <[hidden email]> wrote:
> Hey Guys,
>
> I would like to get acquainted again with compiling QuantLib and adding some
> Swig Python functionality (picking up where I left off a few months ago with
> lots of trial and error - be prepared for questions...).
>
> However I am away from my desktop for the next month and due to home
> networking issues thought I might try my hand at using a Google Compute VM.
>
> Any recommendations for the best setup?
>
> https://cloud.google.com/products/compute-engine/#pricing
>
> Essentially, should I expect memory or cpu bottlenecks?
>
> Thanks!
>
> John
>

> ------------------------------------------------------------------------------
> HPCC Systems Open Source Big Data Platform from LexisNexis Risk Solutions
> Find What Matters Most in Your Big Data with HPCC Systems
> Open Source. Fast. Scalable. Simple. Ideal for Dirty Data.
> Leverages Graph Analysis for Fast Processing & Easy Data Exploration
> http://www.hpccsystems.com
> _______________________________________________
> QuantLib-users mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/quantlib-users
>

--
<https://implementingquantlib.blogspot.com>
<https://twitter.com/lballabio>