SERVING THE QUANTITATIVE FINANCE COMMUNITY

Search found 2749 matches

by N
April 12th, 2007, 5:44 pm
Forum: Off Topic
Topic: How about RC modeling?
Replies: 62
Views: 80098

How about RC modeling?

<t>QuoteOriginally posted by: CactusManNo RCers? I thought that would be a great hobby quants.I have a few RC planes. My favorite is a 65" P-51 with retracts, 15x10x4 blade prop, 50x65 mm brushless outrunner, 36V A123 Li-Ion and Spektrum 7000 xmit/rec. I also have a small chopper that I fly in the o...
by N
April 12th, 2007, 2:53 pm
Forum: Careers Forum
Topic: Citigroup to cut 17,000 jobs
Replies: 18
Views: 75383

Citigroup to cut 17,000 jobs

<t>It's interesting that Citibank mentioned moving backoffice positions to Poland. About five years ago they shifted much of their backoffice to India. I don't think India worked out to be a competitive advantage for them, and that's why they're now exploring Poland.I believe Citibanks biggest probl...
by N
April 12th, 2007, 1:30 pm
Forum: Numerical Methods Forum
Topic: Transformation of Elliptic PDEs to Laplace's Equation
Replies: 63
Views: 100424

Transformation of Elliptic PDEs to Laplace's Equation

<t>QuoteOriginally posted by: CuchulainnQuoteIn the real world, few things are described by linear PDEs. You could also argue that the real world cannot be modelled by any kind of PDE. But that's not the question.I'd argue that the 'real world' is governed by the Hamiltonian and the Hamiltonian can ...
by N
April 12th, 2007, 1:02 pm
Forum: Numerical Methods Forum
Topic: Transformation of Elliptic PDEs to Laplace's Equation
Replies: 63
Views: 100424

Transformation of Elliptic PDEs to Laplace's Equation

<t>QuoteOriginally posted by: ZmeiGorynychClassical mechanics? Where?I'd say everywhere. The Calabi-Yau Manifold (should actually be called the Clifford-Hamilton Manifold) is by far the most common Hamiltonian flow. In the real world, few things are described by linear PDEs. It was well know to math...
by N
April 11th, 2007, 11:14 pm
Forum: Numerical Methods Forum
Topic: Transformation of Elliptic PDEs to Laplace's Equation
Replies: 63
Views: 100424

Transformation of Elliptic PDEs to Laplace's Equation

TJ,Yes, sp(2,C) manifolds (ie, complex manfolds) are quite interesting especially since they appear in classical mechanics and string theory but not in QM. What's really neat is that they are torsion free - the Jacobian is zero.N
by N
April 11th, 2007, 7:36 pm
Forum: Trading Forum
Topic: Developing automated trading systems
Replies: 4
Views: 75699

Developing automated trading systems

A couple of years ago, I did automated FX trading with Oanda. Everything was written in C++. Is Oanda still an FX only shop?
by N
April 10th, 2007, 11:38 pm
Forum: Off Topic
Topic: AMD or Intel?
Replies: 20
Views: 76379

AMD or Intel?

<t>QuoteOriginally posted by: staliSo if I understand this correctly... AMD expects ACML end users to write an interface in assembly to benefit from its ACML_MV vector math lib which for some reason cannot be called from high level languages like Fortran/C.Afaik Intel's VML (and even IBMs mass/SUNs ...
by N
April 10th, 2007, 2:24 pm
Forum: Off Topic
Topic: AMD or Intel?
Replies: 20
Views: 76379

AMD or Intel?

<t>Notes on ACML assembly access for knowledgeable users.As zeta points out the most important thing is to have big shared (among CPUs) L2 cache. So far only the newest Intel Core Duo 2 and AMD X2 processors meet that criteria - but those dual core processors are fairly cheap. AMD will have the firs...
by N
April 6th, 2007, 8:23 pm
Forum: Off Topic
Topic: Global Warming - Impacts, Vulnerability, Adaptation
Replies: 792
Views: 218330

Global Warming - Impacts, Vulnerability, Adaptation

Almost everyone believes the earth is warming (except farmer and pp). The real question is whether it's good or bad. I live in New England and I think it's pretty good - pile on those CO2 emissions! Others who have ocean front property may not be as sanguine.
by N
April 6th, 2007, 3:37 pm
Forum: Off Topic
Topic: AMD or Intel?
Replies: 20
Views: 76379

AMD or Intel?

I also say this elsewhere - to really take advantage of acml, it's best to write the low level interface routines in assembly language so that you fit the code/data into the level two cache (may be .25, .5, 1, 2, or 4 MB depending on CPU).
by N
April 6th, 2007, 2:19 pm
Forum: Off Topic
Topic: AMD or Intel?
Replies: 20
Views: 76379

AMD or Intel?

<t>I think you'll find (in the current competitive CPU market) that Intel and AMD processors that cost the same have about the same performance. I select Intel vs AMD based on the cost of entire system... Right now, I'd give the edge to AMD.Talking about older processors - the P4 was not a bright sp...
by N
April 5th, 2007, 12:42 am
Forum: Numerical Methods Forum
Topic: N's trash talk thread
Replies: 39
Views: 77636

N's trash talk thread

<t>QuoteOriginally posted by: zetaer... correct me if I'm wrong, but one can integrate meromorphic functions via residues/contour integration. Also RNG's based on feedback shift registers are defined over finite fields, not so a meromorphic function which is defined on big ZBoth are division algebra...
by N
April 4th, 2007, 10:21 pm
Forum: Numerical Methods Forum
Topic: N's trash talk thread
Replies: 39
Views: 77636

N's trash talk thread

<t>z,As an example of functions that can't be integrated I was actually thinking of meromorphic functions (I said l-functions, which are related to the geometry of the meromorphic functions and aren't good examples). The RNG is also generated by polynomial division like the meromorphic functions. </t>
by N
April 2nd, 2007, 1:14 pm
Forum: Numerical Methods Forum
Topic: N's trash talk thread
Replies: 39
Views: 77636

N's trash talk thread

A simple example of non-trivial is a multi-valued payoff function. For example f(x)=tan(x) is okay, but f(x)=arctan(x) is not. Another example is when the random number generator output is orthogonal to what you're integrating.
by N
April 2nd, 2007, 10:38 am
Forum: Numerical Methods Forum
Topic: N's trash talk thread
Replies: 39
Views: 77636

N's trash talk thread

low discrepancy sequences are far superiors than true random sequences for integration. True, but low discrepancy sequences/RNGs are still limited to only trivial integration problems.
GZIP: On