Part1. Vectorization of loads/stores group.
GCC 5.0 significantly improves vector code quality for load groups and store groups. By loads/stores group I mean iterated consecutive sequence of loads/stores. For example:
x = a[i], y = a[i + 1], z = a[i + 2] iterated by “i” is loads group of size 3
Group size is distance between smallest and largest loads/stores addresses. In the example (i + 2) – (i) + 1 = 3
This week, we have a many updates and bug fixes to Meshcentral. Under the covers we made significant bug fixes, including to some fixes to quite high priority bugs. Thank you to everyone that keeps submitting feedback and bug reports, it’s very much appreciated. In this mail we will cover three major improvements:
Dear Intel experts,
I wonder if there is any way to control how OpenMP tasks are scheduled by setting some runtime variables? I mean is there any option to be used such as the OMP_SCHEDULE env variable with loop scheduling. Thanks in advance.
In the past week or so, my co-workers and myself have been doing a lot of bug fixing and getting ready to release a bunch of new features. So, in general, working is moving along very well. I released a bunch of server updates yesterday on the info.meshcentral.com web site including a new version of the server installer.
Download Intel XDK and sign up for an Intel XDK account > find this app under Start A New Project > Work with a Demo or find the source code for this sample app here: https://github.com/gomobile/sample-mashery-foodessentials
The mesh team has been hard at work and this week it’s Rick Edgecombe’s turn to shine with his release of the latest Mesh Agent for Android on the Google Play store. The improvements touch many aspects of the Android agent and so, we will have to look at this in detail: