ADSC co-organized GPU technology workshop
In line with its mission to promote high scientific and economic impact of cutting-edge technologies, ADSC collaborated with NVIDIA to help co-organize the GPU Technology Workshop (GTW) South East Asia 2014, which took place in Singapore on July 10. GTW continues the tradition of delivering great content for domain scientists, researchers, industry professionals and developers in Southeast Asia, building onto its big sister, the GPU Technology Conference, which gathers hundreds of GPU computing experts each year.
“Co-organizing NVIDIA’s largest GPU computing event in Southeast Asia developed naturally after a very positive collaboration with them. We welcomed the opportunity of organizing such an event, as it represents an integral part of our core philosophy,” ADSC Director Doug Jones said.
The conference featured a line-up of local and international keynote speakers, four parallel sessions, a dozen cutting-edge technology demos and an exhibition floor showcasing the latest hardware and software applications, representing the context for both beginners and aficionados of the GPU technologies to interact.
The agenda was balanced between morning plenary keynotes featuring Marc Hamilton, vice president of Solution Architecture and Engineering at NVIDIA and John Taylor, director of eResearch and Computational and Simulation Sciences at CSIRO, as well as afternoon sessions covering topics such as high-performance computing and big data, graphics virtualization and VDI, finance and professional visualization and design. One of the afternoon highlights was a keynote address by Cristina Beldica, associate director of the National Center for Supercomputing Applications (NCSA) at University of Illinois at Urbana Champaign.
Additionally, ADSC co-organized a joint seminar on July 9 with A*STAR Computational Resource Centre featuring Beldica to explore NCSA's Blue Waters, a Petascale machine, and the fastest supercomputer on the University of Illinois campus and the challenges of scaling applications on tens of thousands of nodes. A similar presentation featuring A*STAR Computational Centre facilities was presented by Dr. Marek Michalewicz, including a guided tour of their machines.
Post GTW, on July 11, Gabriel Noaje, senior research engineer at ADSC in the Next Generation Compilers and Architectures for Compute Acceleration project team, held a CUDA and OpenACC hands-on workshop in collaboration with NVIDIA. He provided an overview of the three main vectors to leverage the computing power of Graphic Processing Unit (GPU): CUDA programming language, accelerated libraries, and OpenACC directives driven programming. The last session introduced the new Tegra K1 fully programmable embedded chip, and provided hands-on experience using Jetson K1 board, which is already used internally at ADSC for different projects in the fields of social robotics and speech recognition.
ADSC’s goals include contributing to the forward-thinking research climate in Singapore by working with industry partners to further develop ADSC’s research. To continue that, ADSC collaborations with NVIDIA began in November 2011 when a non-disclosure agreement was signed giving ADSC researchers access to confidential information on both hardware and software allowing them to stay on the leading edge of technology breakthrough.
In 2013, NVIDIA and ADSC furthered their collaboration by putting in place the Singapore GPU Users Group. The group is an informal special interest group founded to bring together GPU users from all fields and experience levels in Singapore to learn and share their ideas and creations through informal meetings. The group has regular bi-monthly meet-ups in Fusionopolis organized by ADSC’s Gabriel Noaje and Oxana Plis of NVIDIA. Application developers, researchers, academia, industry and all adopters of GPU computing are welcome to join to share best practices, get feedback on projects and socialize with colleagues and peers across industries.