Intel Speakerships at SC18: Intel Speakerships at SC18 Dallas, TX, United States, November 11-15, 2018

*Abstract submissions are now closed.*

Intel® speakerships at SC18 will feature industry luminaries sharing best practices and techniques to help realize the potential of HPC, HPDA, and AI technologies. Attendees will gain hands-on experience with Intel platforms, hear from Intel® architecture experts, connect with HPC and AI industry leaders, and gain insight on recent technology advances to maximize software efficiency that help drive discovery. Our Technical Committee is looking for speakers to share real science results at the Intel SC18 booth and Intel ancillary events during SC18 in Dallas, Texas.

We are looking for abstracts in the following areas:

1. Technical Talks (20-25 minutes): conveying your techniques
2. Hands-On Tutorials (90 minutes): providing instructions on key steps taken in your discovery

We encourage you to submit abstracts associated with the key topics listed below.

Artificial Intelligence: Machine Learning/Deep Learning: Machine Learning is transforming everything from self-driving cars to health care analytics. We welcome deep dive technical sessions that focus on machine learning research, application and scaling techniques in HPC environments. Content that addresses: ML Optimization or Results on Intel Platforms, Storage, File Systems, Databases

Cloud: HPC in the cloud by measures of Hyperion Research* has nearly 2/3 of the HPC sites they are in touch with running some of their jobs in public clouds but of those sites, they run less than 10% of all of their jobs. There are still challenges a plenty with data movement, performance, networking, storage, security, ROI, etc.  We welcome technical sessions that show the progress, successes, setbacks and gotcha’s around using “HPC in the cloud” whether that cloud is public or private or hybrid. Content that addresses: HPC in the cloud          _{*Hyperion Research – ISC2018 Report out}

High Performance Data Analytics: This area focuses on how real-world enterprises harness the power of Intel’s latest platform in HPC, AI, and Data Analytics to drive business growth and innovation. Artificial intelligence is increasingly used within new generation modeling and simulation workloads along with a rise in hybrid workloads. The new generation HPC infrastructure should cater to HPC, AI, and Data Analytics workloads all using one infrastructure (i.e. unified architecture). Content that addresses: High Performance Data Analytics (HPDA), HPC and AI convergence, Unified Architectures, Machine Learning, Enterprise Integration, Spark on HPC, Containers in HPC, Hybrid Cloud in HPC, SLURM and MESOS Integration

High Productivity Languages: “Good, Cheap, and Fast… Pick Two” is the adage that is often used to highlight trade-offs in selecting qualities that you desire. In programming, scripting languages typically fall under the “faster to learn, easy to write but not performant” where the two qualities are the reasons why you picked it in the first place, but you would like all three! This area will focus on how these high productivity languages such as Python, R, and Julia are helping drive not only emerging research in HPC but also being deployed at scale in Big Data/Machine Learning HPC environments. Content that addresses: Python, R, MATLAB, Julia

Parallel Programming (HPC): Parallelism is critical to achieving performance at all levels of the computing landscape, from small edge devices to the largest supercomputers. Taking advantage of parallelism requires the right programming tools, techniques, and knowledge. We welcome experiences in all aspects of code modernization from vectorization to threading to message passing on a variety of parallel computing platforms. Content that addresses: Parallel Algorithms, Parallel Programming Tools or Libraries, Parallel Programming Debugging, Parallel Programming Optimization Techniques

Storage (AEP, NVMe, DAOS): HPC has always been the domain of HUGE compute capacity in flops but to feed the computational monster HUGE amounts of data are required and at be at the ready. HPC, Big Data, and AI are all data intensive in the ways of volume, velocity, and variety, and thus storage of this data and the ability to deliver it as needed is a never ending challenge. For the areas of storage we look for technical sessions that focus on meeting these challenges at the application, system, and I/O level. Content that addresses:  Lustre, DAOS, SSD/Persistent Memory Storage, Parallel File Systems

Systems: Configuration, Management & Cloud – Standing up HPC systems, keeping them healthy and running optimally is challenging at any scale. Additionally, the availability of cloud based HPC resources and containerization has fostered new usage models. This area’s focus will be on the deployment and management of HPC resources along with their complex software ecosystems. Content that addresses: OpenHPC, System Configuration and Management, Cloud (setup or usage), and Portability

Visualization Development: The latest highly parallel Intel® processors enable high performance, interactive and high fidelity (up to photo-realistic) CPU based rendering using libraries such as Mesa with OpenSWR, Embree, and OSPRay. We welcome case studies, implementations, and research involving CPU based Visualization applications and infrastructure. Content that addresses: OpenSWR, OSPRay, Embree, ParaView, VTK, VTK-m, VisIt, VMD, EnSight, In-Situ Visualization

Speaker Deadlines (updated!)

Venue

Intel SC18 Booth #3223 and One Intel Station (located at Union Station, 400 South Houston St, Dallas, only a 5 - 7 minute walk from SC18.

Contact

All questions about speaker submissions should be emailed to: ipcc.program.office@intel.com attn: SC’18 Intel Speakerships.