Live presentations

Join Xilinx and partners at SC21 as we explore the latest advances in high performance computing.

*Live presentations from our virtual booth (Central Time)

How Heterogeneous Computing is Driving a New Era of Accelerated HPC Workloads

Date: Nov. 15, Monday
Time: 12:15 PM (CST)

The datacenter landscape has changed dramatically in the last decade and one thing is clear: CPUs and GPUs alone can no longer meet the ever-growing demand of big-data-driven workloads that are pushing toward the exascale threshold.

But which computing architectures will help you meet the right mix of price, performance and power efficiency? Learn how compute-intensive workloads such as computer-aided engineering, massive streaming sensor data, and accelerated graph database analytics can benefit from new FPGA-based HPC clustering methodologies to deliver new levels of innovation and understanding. 


Accelerating LS-DYNA with Xilinx Alveo™

Date: Nov. 16, Tuesday
Time: 11:30 AM (CST)

Learn how LS-DYNA from ANSYS is utilizing Xilinx Alveo accelerator cards to accelerate real-world simulations of complex real-world problems used by au­to­mo­bile, aero­space, con­struc­tion, mil­i­tary, man­u­fac­tur­ing, and bio­engi­neer­ing in­dus­tries. 


BlackLynx – Accelerating Spark Queries 10x with Xilinx

Date: Nov. 16, Tuesday
Time: 2:45 PM (CST)

Datasets have grown beyond the capability of CPUs to index and search. The classic method of ETL (Extract, Transform, Load) requires too much processing and is too costly as datasets approach the 1PB or more range. In this video we demonstrate how Xilinx can accelerate the search of unindexed data as a prefilter for the well-known ETL query engine Splunk using a common unindexed dataset – network PCAP data. We will show performance results that typically exceed 100x improvement over traditional Splunk queries and how you can perform more cost effective queries on your large datasets.


Improving Spark Efficiency with NoLoad Transparent Compression from Eideticom, Cisco, and Xilinx

Date: Nov. 17, Wednesday
Time: 12:15 PM (CST)

Learn about accelerated compression and optimization of storage in big data analytics with NoLoad® FPGA-based acceleration. Cisco, Xilinx, and Eideticom bring together a high-performance and scalable heterogeneous system architecture that delivers improved storage efficiency by achieving better compression and optimization of data to reduce TCO.

Demos and Presentations on-demand

In addition, don’t miss the following demos and presentations on-demand

Customizable FPGA Cloud Solutions with VMAccel and Xilinx

VMAccel is an FPGA-accelerated cloud computing platform that provides cost-effective, customizable, and portable solutions for complex HPC workloads. VMAccel’s FPGA-as-a-service infrastructure enables users from anywhere in the world, with any skill level to implement FPGAs into their computational workloads with ease via the cloud.


Plipos – Enabling Fast and Cost Effective Data Storage

Learn how Pliops is radically simplifying how data is processed and SSD storage is managed. Learn how Pliops and Xilinx are working together to deliver faster and more cost effective data storage. 


Snowlake – Accelerating Molecular Dynamics Beyond GPUs

Learn how Snowlake is dramatically accelerating dynamic simulations of bio-molecules with Xilinx Alveo and achieving higher performance over CPU and GPUs.


TigerGraph Cosine Similarity Acceleration for Healthcare Provider and Patient Similarity Analytics

TigerGraph, provides graph analytics software, accelerated with Xilinx Alveo cards and based on the massively parallel processing capability of FPGA architecture, offers superior results for computing cosine similarity calculations. Cosine similarity is used in data mining for measuring the similarity of data objects within clusters. 


TigerGraph Louvain Modularity Acceleration for Fighting Financial Crimes

Xilinx has partnered with TigerGraph for accelerating graph analytics with one use case being fighting financial crime in financial services. TigerGraph Financial Crimes Toolkit helps enterprises detect and combat fraudsters, money launders and other financial criminals faster and more accurately. TigerGraph uses Xilinx to accelerate graph algorithms such as Louvain Modularity for Community Detection. Customers see speed-up and cost savings using the latest Alveo cards when compared to traditional CPU technology. 

Technical Track Workshop

Seventh International Workshop on Heterogeneous High-Performance Reconfigurable Computing

Date: Nov. 15, Monday
Time: 9:00 AM (CST)
Speaker: Ken O'Brien, Senior Research Scientist, Xilinx


ACCL: FPGA-Acceleration Collectives Over 100Gbps TCP-IP

Date: Nov. 15, Monday
Time: 10:30 AM (CST)
Speaker: Michaela Blott, Xilinx Principal Engineer 


Panel Session - SmartNICs: The Future or Folly?

Date: Nov. 19, Friday
Time: 8:30 AM (CST)
Speaker: Dr Gordon Brebner, XIlinx Fellow