KTM Innovation develops novel use-cases for enhancing the drivers' safety and experience by continually analyzing driving related data. With @elegant_ict data processing will become faster and more secure via the unified ELEGANT API.
#KTM#ReadyToRace@EU_H2020#H2020🇪🇺
We just open-sourced a new Java JNI library for GPU programming for Intel Level Zero devices.
This first release supports a subset of LevelZero 1.1.2 and can dispatch the execution of SPIRV kernels
#oneAPI#levelZero#JNI#Java#GPUProgramming
https://t.co/emJR0BMVtc
Despite the fact that #JavaZone in-person conference was cancelled, it is very much appreciated the organizers are making a huge effort for making this happen
On Thursday, I will be talking about our research: Transparent Hardware Acceleration for Java
https://t.co/GssR82hZA2
Major release for #TornadoVM 0.12 with support for #SPIRV code generation and integration with #Intel LevelZero 🎉
Check out the latest features and improvements: https://t.co/hqUfSHgtHB
Docker images: https://t.co/D7fPM9DlHU
#oneAPI#java#GPU#FPGA@elegant_ict
Excited to be speaking at the @QCon Plus online software development conference this November.
We'll deep-dive into the #TornadoVM APIs for GPU and FPGA programming, and give an overview of the OpenCL and PTX backends!
More information on the talk:
https://t.co/OccdfWe351
Great article by Juan Fumero describing in detail with hands on examples how to perform fine grain timing of #GPU accelerated code with #Intel#LevelZero
Finally, I took the time to clean up a recent project: #ProtonVM, a Parallel Bytecode Interpreter For Heterogeneous Hardware.
ProtonVM runs a simple and very small bytecode interpreter on #GPUs and #FPGAs!
https://t.co/exy16wf1Ci
And, it was presented @MoreVMs 2020!
Our latest blog about #TornadoVM and how it will complement the @elegant_ict architecture for accelerating arbitrary code on heterogeneous #IoT#Cloud deployments - acceleration-as-a-service on the horizon!
https://t.co/ZEwqMv4enJ
One of the goals of OneAPI design was “No developer left behind” - level zero is a key layer that enables other api, language and tool developers to improvise without the constraints of higher level abstractions.
https://t.co/HDXmb8W9Pz