Today’s high-performance computing (HPC) systems are unleashing discovery and insights at an unprecedented pace. The intersection of artificial intelligence (AI) and HPC is transforming industries from healthcare to transportation while solving some of the toughest socio-economic and scientific challenges in our world. Intel offers a hardware and enabling software portfolio of unprecedented choice and scale to power this next wave of discovery. To highlight the convergence of HPC and AI, Intel is pleased to be sharing seven technical papers – one of them a Best Paper finalist – as well as five conference tutorials, five workshops, one panel, and two posters.
For a look at exciting examples of how organizations around the globe are using Intel technology to power their AI applications, plus the opportunity to take home some BunnyPeople™collectible pins, come visit us at Booth #3223. For even more demos and presentations, along with a series of AI-focused hands-on tutorials, check out our off-site One Intel Station at the historic Dallas Union Station.
Demos in Intel Booth #3223:
GE Healthcare: CT Scanner Technology Powered by Intel AI: A collaboration with GE Healthcare, this demo shows Intel is bringing the power of AI to clinical diagnostic scanning and other healthcare workflows. Featured Intel technologies: Intel® Xeon® Scalable processors, Intel optimizations for TensorFlow*, and deep learning tools included in the Intel® Distribution of OpenVINO™ toolkit
Novartis: nClassification of high content Cellular Images on Intel AI: See scale out training of a multi-scale CNN-based high content imaging (HCI) classifier on 330K images in under an hour. Featured Intel technologies: Intel Xeon Scalable processors, Intel® Omni-Path Architecture (Intel® OPA), Intel® Optimization for TensorFlow*, Intel® Math Kernel Library for Deep Neural Networks (Intel® MKL-DNN)
RocketML*: A Convergence Platform for HPC and AI: RocketML, a hardware and software co-optimized platform tuned for rapid experimentation, shows real-time training on any data size on CPUs. Featured Intel technologies: Intel Xeon Scalable processors, Intel® Math Kernel Library
Inspur: FPGA Accelerator Empowers Object Detection Application: This live demo shows how the Intel-based Inspur F10A FPGA Accelerator and TF2 Compute Acceleration Engine, supporting the TensorFlow* machine learning frAMework, delivers high performance and low latency for AI applications. Featured Intel technology: Intel® Arria® 10 GX 1150 FPGA Device.
You’ll find us making contributions across SC18 in paper presentations, tutorials, workshops, panels, and talks.
- Extreme Scale De Novo Metagenome Assembly Best Paper Finalist Presentation: Tuesday, November 13, 1:30 - 2 PM, Location C140/142
- Optimizing High Performance Distributed Memory Parallel Hash Tables for DNA k-mer Counting Presentation: Tuesday, November 13, 2 - 2:30 PM, Location C140/142
- Many-Core Graph Workload Analysis Presentation: Tuesday, November 13, 3:30 - 4 PM, Location C146
- Framework for Scalable Intra-Node Collective Operations Using Shared Memory Presentation: Wednesday, November 14, 11 - 11:30 AM, Location C140/142
- High-Performance Dense Tucker Decomposition on GPU Clusters Presentation: Wednesday, November 14, 2:30 - 3 PM, Location C140/142
- CosmoFlow: Using Deep Learning to Learn the Universe at Scale Presentation: Thursday, November 15, 2 - 2:30 PM, Location C140/142
- Anatomy of High-Performance Deep Learning Convolutions on SIMD Architectures Presentation: Thursday, November 15, 2:30 - 3 PM, Location C140/142
- OpenMP* Common Core: A “Hands-On” Exploration Sunday, November 11, 8:30 AM - 5 PM, Location C145
- Advanced OpenMP: Host Performance and 5.0 Features Sunday, November 11 8:30 AM - 5 PM, Location C141
- Programming Your GPU with OpenMP: A Hands-On Introduction Monday, November 12, 8:30 AM - 5 PM, Location C142
- Mastering Tasking with OpenMP Monday, November 12, 8:30 AM - 12 PM, Location C141
- Performance Tuning of Scientific Codes with the Roofline Model Monday, November 12, 1:30 - 5 PM, Location C154
- Training Speech Recognition Models on HPC Infrastructure Sunday, November 11, 5 - 5:30 PM, Location D167/174
- Learning to Lead in HPC - Strategies to Start Your Leadership Journey Monday, November 12, 9:30 - 10 AM, Location A302
- Ninth Annual Workshop for the Energy Efficient HPC Working Group (EE HPC WG) Monday, November 12 9 - 9:15 AM, Location D221
- Large Minibatch Training on Supercomputers with Improved Accuracy and Reduced Time to Train Monday, November 12, 5 - 5:30 PM, Location D167/174
- International Workshop on Performance, Portability, and Productivity in HPC (P3HPC) Friday, November 16, 8:30 - 8:40 AM, Location D173
- Swiss Army Programming: Performance and Portability from Modern Tools Tuesday, November 13, 1:30 - 3 PM, Location C147/148/154
- Tensorfolding: Improving Convolutional Neural Network Performance with Fused Microkernels Thursday, November 15, 8:30 AM - 5 PM Location C2/3/4 Ballroom
- Tensor-Optimized Hardware Accelerates Fused Discontinuous Galerkin Simulations Thursday, November 15, 8:30 AM - 5 PM, Location C2/3/4 Ballroom
On Wednesday, November 14, Vikram Saletore, Intel principal engineer, will be giving a talk on the convergence of HPC, data analytics, & AI at the Lenovo booth #3617 from 2 - 3 PM.
You’ll find us outside the hall, too – at One Intel® Station
During SC18, Dallas Union Station at 400 S. Houston Street is One Intel Station. Just a short walk from the Hutchinson Convention Center, it’s the place to experience AI theory in action. You can see real-world use cases from Intel® AI Builders members, attend tutorials, and more. Although seats for our AI Crash Course on Wednesday, November 14, 2:00PM-5: 30 PM are already booked, One Intel Station offers a wide array of events and demos in the “AI Zone” located in the Station’s Grand Hall. We’ll provide snacks and drinks in the reCharge Lounge, and you’ll have a chance to talk with our HPC and AI subject experts on the latest trends and implementations in AI.
Demos at One Intel Station
- Using Intel Optimization for TensorFlow* to Accelerate Unity ML-Agents* training
- Training a Chatbot Using Intel Optimization for TensorFlow*
- Multi-threaded Reinforcement Learning
- Parallelized Word2Vec on Intel Xeon Processors
Hands-on Tutorials at One Intel Station
- An Introduction to OpenHPC*: Monday, November 12 from 12-2:30 PM
- Compiling and Tuning for Performance Using Intel® AVX-512: Monday, November 12 from 4–5:30PM
- HPC+AI Convergence: Running Machine Learning & Deep Learning along with Big Data Analytics in HPC Environments: Tuesday, November 13 from 12-1:30 PM
- Performance Tuning with Cache-aware Roofline Model in Intel® Advisor: Tuesday, November 13 from 4–5:30 PM
- Parallel Fortran* 2018: Achieving Asynchronous Execution with Events and Visualizing Performance with TAU*: Wednesday, November 14 from 12-1:30 PM
Posters from Intel AI Student Ambassadors at One Intel Station
The Intel Student Ambassador Program, part of the Intel AI Academy, works with Graduate & Ph.D. students at top academic institutions around the world to drive innovative new AI research and use cases across numerous verticals and industries. Selected members of the program have been chosen to present their work:
- RAIL - Risk Averse Imitation Learning - Indian Institute of Technology, Kharagpur
- Explaining the Unexplained: A Class-Enhanced Attentive Response (CLEAR) Approach to Understanding Deep Neural Networks - University of Waterloo/Vector Institute Toronto
- Towards Autonomous UAV Flight in Cluttered Forestry Environments - DurhAM University, UK
- Deep Learning for Cryptocurrency Trading - Rutgers University
- Where’s the Bear? - Automating Wildlife Image Processing Using IoT and Edge Cloud Systems - University of California Santa Barbara
- Self-Driving Vehicles and Risk Analysis - University of Southern California Virtebai School of Engineering
Follow Intel for ongoing news about AI advances
As AI is applied to ever larger and more complex data sets, Intel delivers an ever advanced foundation for HPC, accelerating the path to deeper discovery and rewarding insights.
For more about Intel Xeon Scalable processors and other technologies to address the complex spectrum of AI and HPC workload requirements visit https://ai.intel.com.
Intel technologies’ features and benefits depend on system configuration and may require enabled hardware, software or service activation. Performance varies depending on system configuration. No computer system can be absolutely secure. Check with your system manufacturer or retailer or learn more at intel.com.
Intel, the Intel logo, OpenVINO, and Xeon are trademarks of Intel Corporation or its subsidiaries in the U.S. and/or other countries.
*Other names and brands may be claimed as the property of others.
© Intel Corporation