- Address:Off Sunyani-Fiapre Road, Ghana
UENR ORGANIZES FIRST EVER NVIDIA DLI EVENT IN AFRICA
A workshop on Fundamentals of Accelerated Computing organized by the NVIDIA Deep Learning Institute (DLI) and the University of Energy & Natural Resources (UENR) has ended in Sunyani. This DLI event was the first of its kind to be organized in Africa.
The two (2) -day workshop which was facilitated by the Earth Observation Research & Innovation Centre (EORIC), a research arm of the University, offered hands-on training for developers, data scientist and researchers looking to solve challenging problems with deep learning and accelerated computing.
The workshop took place on the 28th and 29th of October, 2019. Participants included lecturers, researchers, teaching assistants and students from the University of Energy & Natural Resources (UENR) and the Kwame Nkrumah University of Science & Technology (KNUST) in Kumasi.
Specifically, the workshop sought to teach participants the fundamental tools and techniques for accelerating C/C++ applications to run on massively parallel GPUs with CUDA®. Participants also practised writing code, configuring code parallelization with CUDA, optimizing memory migration between the CPU and GPU accelerator, and implementing the workflow that was learnt on a new task.
The workshop was outlined as follows;
| Topic | Description |
| Accelerating Applications with CUDA C/C++ | Learn the essential syntax and concepts to be able to write GPU-enabled C/C++ applications with CUDA: > Write, compile, and run GPU code. > Control parallel thread hierarchy. > Allocate and free memory for the GPU. |
| Managing Accelerated Application Memory with CUDA C/C++ | Learn the command line profiler and CUDA managed memory, focusing on observation-driven application improvements and a deep understanding of managed memory behaviour: > Profile CUDA code with the command line profiler. > Go deep on unified memory. > Optimize unified memory management. |
| Asynchronous Streaming and Visual Profiling for Accelerated Applications with CUDA C/C++ | Identify opportunities for improved memory management and instruction-level parallelism: > Profile CUDA code with the NVIDIA Visual Profiler. > Use concurrent CUDA streams. |
Mr. Alexander Tsado, the Technical Product Manager (External), Cloud Services GPU Business at NVIDIA and Founding Director of Alliace4AI, interacted with the participants through Skype on the second day of the workshop. He expressed his excitement about certain African based projects and initiatives that applied Artificial Intelligence (AI) and Deep Learning techniques to solve problems in Africa.
Mr. Tsado then encouraged the participants to take the course seriously and expressed believe in participants coming up with ground-breaking technologies in the future.
Dr. Mark Amo-Boateng, head of Earth Observation Research & Innovation Centre (EORIC) and lecturer at the Department of Civil and Environmental Engineering in the University of Energy & Natural Resources (UENR) was the lead facilitator. He was supported by Nana Ekow Nkwa Sey, who served as the Teaching Assistant for the training.
Dr. Amo-Boateng explained the basic concepts of accelerated computing, demonstrated how to apply these concepts to optimize C/C++ codes to run on the GPU and discussed a number of real-life application areas where such technologies are currently being applied.
At the end of the workshop, participants were given access to additional resources to create new GPU-accelerated applications. Participants were encouraged to complete a GPU task and successfully execute it in order to obtain a certificate.
Beneficiaries of the workshop expressed their appreciation and satisfaction at the content of the training and requested that similar training be organized to improve their knowledge in the application of artificial intelligence, deep learning and accelerated computing.
