GPU computing is a vast resource that is still, in great part, untapped, as software does not have the code necessary to recognize and use the parallel processing might of NVIDIA CUDA cores and AMD stream processors.
Advanced Micro Devices
isn't about to let this situation persist any longer than it has to. In fact, the company has formed a team with Oracle, so as to enable heterogeneous computing in Java.
This would make the software compatible and capable of running at optimum efficiency on a wider variety of servers and cloud data centers.
At the same time, it will increase the relevance of AMD FirePro cards and FireStream GPU accelerators.
After all, if a popular, widely used software tool and developer kit like Java gains the ability to use GPUs alongside CPUs or APUs, AMD's products will have a greater demand.
The OpenJDK Project “Sumatra” will try to find out how the Java Virtual Machine (JVM), as well as the Java language and APIs, can be enhanced to allow applications to take advantage of graphics processing unit (GPU) acceleration.
Both discrete graphics cards and high-performance graphics processor cores integrated in (APUs) will be taken into account.
On that note, “Sumatra” will provide developers with mainstream programming models and guidance on enabling support for other JVM-based languages such as Scala, JRuby and Jython.
“We expect our work with AMD and other OpenJDK participants in Project “Sumatra” will eventually help provide Java developers with the ability to quickly leverage GPU acceleration for better performance,” said Georges Saab, vice president, software development, Java Platform Group at Oracle.
"We hope individuals and other organizations interested in this exciting development will follow AMD's lead by joining us in Project “Sumatra."
This has to be one of the projects that AMD isn't about to dump any time soon, even though it is slightly drowning in bureaucracy
at the moment.