Nowadays GPU resources are available not only on gaming PCs, but also on laptops or cloud machines. We can utilize them to get huge optimizations on a commodity hardware.
Dmitriy firstly used GPU while working on a fintech startup that makes a lot of computations. At first there were used regular methods of parallel data processing using CPU, but later it became clear that GPU is a powerful resource for such tasks that was unused. After some research a working model was made and computations started using both CPU and GPU.
This talk will start with a brief introduction focusing on GPU evolution and then turn to using GPU in Java. We’ll briefly meet OpenGL and then go on to examine CUDA and OpenCL — technologies of massive general-purpose computation on GPU. We’ll see what are the pros and cons of using GPU in Java (and programming in general). We'll explore how to automate GPU using in Java and run it all in the cloud.
Dmitry Aleksandrov, T-Systemsbercut2000
Dmitry is a principal expert developer at T-Systems. He has more than a decade experience mainly in Java Enterprise in banking/telecom, but interested in dynamic languages on JVM and features like massive computations on GPUs. He is a Java Champion, a true believer in open source and community driven initiatives. He is a co-lead of the Bulgarian Java User Group and co-organizer of jPrime Conf. Dmitry is a blogger and also a frequent speaker at local events as well as conferences like JavaOne, Devoxx/Voxxed and Joker.