[MCX Newsletters] MCX v2016.4 is released!

[Qianqian Fang]

Dear MCX and MMC registered users,

It is my great pleasure to announce that the new MCX release,
v2016.4 (aka 1.0-beta), is now available for download. The
Release Note and download link can be found at the MCX's newly
redesigned website:

http://mcx.space

As a registered user, you can directly access the updated packages,
including both MCX and MCXLAB, at

https://sourceforge.net/projects/mcx/files/

MCX v2016.4 is a major milestone release (probably I should
have called it MCX 2.0 :) ) and has accumulated
numerous improvements over the past two years. It
contains a list of major upgrades of new features
and fixes of critical bugs. If you are using an older release of
MCX, you are strongly recommended to upgrade immediately.

In this new release, we implemented a fast precise ray-tracing
algorithm in MCX to propagate photons inside a voxelated domain,
replacing the old approximated ray-tracer. This update leads
to significantly improved solution accuracy when a medium
contains scattering contrast from the background. Moreover,
MCX now enables atomic operations by default. This means
more accurate results can be obtained near the source
regions. Another major improvement is the support of
multi-GPUs. One can now simultaneously use multiple
GPUs in a computer to obtain proportional speed improvement!
If you have a Maxwell GPU, the new software has been
optimized to gain 2.5x speedup on these cards.

In addition, the new MCX now uses atomic operations by default
and is equipped with two new random number generators that
are supposed to be both faster and higher quality. If you
want to make changes to MCX, the new source code package
contains Nsight project files for both Eclipse (Linux/Mac) and
Visual Studio (Windows). Compilation becomes easier for those
who prefers IDEs.

There are more enhancements coming with this release, including
a completely rewritten website (http://mcx.space ), GPU Contest
(http://mcx.space/gpubench ) and MCX Speed Challenge. I will
make announcements in a separate email.

I would like to thank Fanny Paravecino and Leiming Yu, the PhD
students who had spent significant effort making MCX faster
on the GPUs, and Ruoyang Yao, who have helped a lot on the
algorithm development. I also would like to thank my collaborators,
Drs. David Kaeli from Northeastern Univ. and Xavier Intes from
RPI who have been actively engaged with this research. My
thanks also extend to many users who have shared their feedback
and suggestions.

This release is not possible with the funding support from the
NIH/NIGMS (R01-GM114365). We cordially thank NIH for the
support.

Enjoy the new software and please let me know if you have any
problems or comments.

Qianqian