InfoQ Homepage Field Programmable Gate Arrays (FPGA) Content on InfoQ
Articles
RSS Feed-
/articles/java-performance-tornadovm/en/smallimage/logo-1646667233753.jpg)
Level up Your Java Performance with TornadoVM
GPUs, FPGAs, or multi-core CPUs are present in almost every computing system today. These devices help increase performance and run more efficient workloads, but most frameworks are built on C or C++ only. At QCon Plus, Juan Fumero spoke about TornadoVM, a high-performance computing platform for the JVM, allowing to offload, at runtime, Java code to run on heterogeneous hardware accelerators.
-
/articles/tornadovm-java-gpu-fpga/en/smallimage/java-1591918279658.jpeg)
TornadoVM: Accelerating Java with GPUs and FPGAs
The proliferation of heterogeneous hardware represents a problem for programming languages such as Java that target CPUs. TornadoVM extends the Graal JIT compiler to take advantage of GPUs & FPGAs and provides a flexible, high-level model whilst still enabling high performance and features such as live task migration.