SIMD Parallel Programming with the Vector API By José Paumard

The first version of the Vector API was published as an incubator feature in the JDK 16. We now have the 6th incubator version in the JDK 21, which is stable enough to take a look at it, and see how we can use it. The Vector API can tremendously speed up your in-memory computations by using the SIMD (Single Instruction Multiple Data) capabilities of the cores of your CPU. The SIMD architecture is not a new concept, as it was already used in parallel computers in the 80s. This session explains the differences between parallel streams and parallel computing, and how SIMD computations are working internaly on simple examples. It then shows the patterns of code that the Vector API is giving along with their performances, and how you can use them to improve your in-memory data processing computations. More advanced techniques are also presented, to go beyond the basic examples.

JOSÉ PAUMARD

[karlbrodowsky]

I have actually programmed on a connection machine....