NECSTFridayTalk - Lowering the barriers to the GPU world

Ian Di Dio Lavore
Master student in Computer Science and Engineering
NECSTLab, Politecnico di Milano
Event will be online from Facebook
May 20th, 2022
1.00 pm
Contacts:
Marco Santambrogio
Research Line:
System architectures
Master student in Computer Science and Engineering
NECSTLab, Politecnico di Milano
Event will be online from Facebook
May 20th, 2022
1.00 pm
Contacts:
Marco Santambrogio
Research Line:
System architectures
Sommario
On May 20th, 2022 at 1.00 pm "Lowering the barriers to the GPU world" a new appointment of NECSTFridayTalk, will be held online via Facebook by Ian Di Dio Lavore, Master student in Computer Science and Engineering at NECSTLab, Politecnico di Milano.
GPUs are readily available in cloud computing and personal devices, but their use for data processing acceleration has been slowed down by their limited integration with common programming languages such as Javascript, Python or Java. Exploiting the full capabilities of multi-GPU systems is an arduous task, due to the complex interconnection topology between available accelerators and the variety of inter-GPU communication patterns exhibited by different workloads. A software engineer needs to be trained in advanced asynchronous programming in order to take advantage of those capabilities.
Therefore, we present a novel scheduler for multi-task GPU computations that provides transparent asynchronous execution on multi-GPU systems without requiring prior information about the program dependencies or the underlying system architecture.
Our scheduler integrates with the polyglot GraalVM ecosystem, and is therefore available for multiple high-level languages, providing a general framework that can significantly lower the barriers to access the multi-GPU acceleration world. We validate our work on a set of benchmarks designed to investigate scalability and inter-GPU communication.
Our scheduler automatically achieves 80-90% peak performance against hand-optimised CUDA host code on Volta and Ampere multi-GPU systems.
GPUs are readily available in cloud computing and personal devices, but their use for data processing acceleration has been slowed down by their limited integration with common programming languages such as Javascript, Python or Java. Exploiting the full capabilities of multi-GPU systems is an arduous task, due to the complex interconnection topology between available accelerators and the variety of inter-GPU communication patterns exhibited by different workloads. A software engineer needs to be trained in advanced asynchronous programming in order to take advantage of those capabilities.
Therefore, we present a novel scheduler for multi-task GPU computations that provides transparent asynchronous execution on multi-GPU systems without requiring prior information about the program dependencies or the underlying system architecture.
Our scheduler integrates with the polyglot GraalVM ecosystem, and is therefore available for multiple high-level languages, providing a general framework that can significantly lower the barriers to access the multi-GPU acceleration world. We validate our work on a set of benchmarks designed to investigate scalability and inter-GPU communication.
Our scheduler automatically achieves 80-90% peak performance against hand-optimised CUDA host code on Volta and Ampere multi-GPU systems.
The NECSTLab is a DEIB laboratory, with different research lines on advanced topics in computing systems: from architectural characteristics, to hardware-software codesign methodologies, to security and dependability issues of complex system architectures.
Every week, the “NECSTFridayTalk” invites researchers, professionals or entrepreneurs to share their work experiences and projects they are implementing in the “Computing Systems”.
Streaming will be available via Facebook