Quant development: What can NAG quant analytics library do for you in 2012!
Happy New Year to you all.
It’s fantastic to see so many new members to the group. The official Numerical Algorithms Group on Linkedin has over 600 members and is growing daily.
I’d like to ask members of the group a question. If NAG could do one thing for you in 2012 what would it be (product or service related requests please :-))? We’ll try to respond to each request.
==
Expand the coverage of your GPU product. One possible ‘quick-win’ would be to implement GPU matrix-matrix multiply and then use this as a drop-in replacement for the standard CPU dgemm in various routines. This is the approach taken by Maple for example.
Thank you Mike. NAG have done some experimental work in this area. I’ll see if I can get an update and find out what information we can share.
Keep up the excellent online documentation. That documentation is a very valuable resource.
Thanks for that. That’s good to hear.
==
good suggestion, and one which we have been actively investigating. Our plans are to make GPU accelerated versions of our CPU libraries available for experimentation and evaluation.
These libraries will make run-time decisions about where all level 3 BLAS operations take place, whether on the CPU or on the GPU (using CUBLAS), based on auto-tuning data. We are also looking into using a selection of important routines from MAGMA in this manner, thus providing some GPU-accelerated LAPACK functionality.
This will enable existing users of the NAG libraries to re-link their applications and make use of GPUs for appropriate problem sizes without changing source code.
We are currently putting CUBLAS and MAGMA through our stringent tests in order to ensure the same quality of results are returned to the user wherever the computation takes place.
==
That’s great, thanks. Sounds like there will be a lot of new technology to try out in 2012–I’m looking forward to it
NOTE I now post my
TRADING ALERTS into my personal
FACEBOOK ACCOUNT and
TWITTER. Don't worry as I don't post stupid cat videos or what I eat!