Friends of OpenJDK Today
Foojay Podcast #17: Execute Java Code with TornadoVM on CPUs, GPUs, and FPGAs
March 27, 2023TornadoVM is a programming and execution framework for offloading and running JVM applications on multi-core CPUs, GPUs, and FPGAs.
With the same code, some of your existing program code can be executed hundreds of times faster!
Guests
- Juan Fumero, TornadoVM Lead Architect
- Christos Kotselidis, TornadoVM Project Leader
- Thanos Stratikopoulos, TornadoVM Senior Solutions Architect
- Jakob Jenkov
Podcast
- Host: Erik Costlow
- Production: Frank Delporte
Content
- 00’00 Intro
- 00’36 Introduction of the guests
- 04’26 What is TornadoVM?
- https://foojay.io/today/hardware-acceleration-for-java-tornadovm-can-do-it/
- https://fosdem.org/2023/schedule/event/hardware/
- https://www.tornadovm.org/
- https://github.com/beehive-lab/TornadoVM
- https://www.infoq.com/articles/java-performance-tornadovm
- https://www.infoq.com/articles/tornadovm-java-gpu-fpga
- 05’54 How applications can make use of the acceleration provided by TornadoVM
- 11’48 The difference between CPU threads and GPU instruction chain
- 13’42 Possible use cases for TornadoVM
- 15’23 Results on Apple M1
- 17’19 Can TornadoVM be used in cloud environments
- 21’18 How to use the API
- 24’41 Jakobs view of what would be a good match between TornadoVM and cloud usage on AWS Lambdas
- 30’54 The complexity of GPU and FPGA programming languages and handling the differences between different architectures of GPUs, CPUs, and FPGAs
- 40’28 How TornadoVM could be used to heat up buildings, help to reduce the total cloud cost for companies, and run ChatGPT
- 43’30 Relationship between project Panama and TornadoVM
- 48’10 How to get started with TornadoVM
- 54’41 Outro