Innovation2018-11-01T23:01:17+00:00

Innovation

Innovation at Falcon

Founded in 2014 by Dr. Jason Cong, Director of the Center for Domain Specific Computing (CDSC) at UCLA, Falcon Computing is an application acceleration company with strong academic roots.

As part of our mission to simplify acceleration on heterogenous hardware platforms, we partner with academic institutions working on cutting-edge computation techniques to solve world-class problems. The flexibility of FPGAs makes them a great tool for research and experimentation which can be simplified with the use of our software products as offered under our special Academic Program pricing.

Academia Partners

UCLA Technology Center for Genomics
& Bioinformatics (TCGB)

  • Fully automated, high-throughput genomic facility equipped with all major sequencing and microarray platforms.

  • Provides services nationwide across the US.
    (http://pathology.ucla.edu/tcgb)

Genomics Partnership with
Peking University

  • Successful evaluation of Falcon Genomics Appliance deployed at the High Performance Computing Platform of the Center for Life Science at Peking University

  • Results validated to be the same with GATK 3.5 using precisionFDA.org dataset and Genotype-Tissue Expression (GTEx) dataset.
  • 50x speedup compared to their existing HPC approach

Academic Conferences & Publications

Featured Publication

In collaboration with UCLA, Georgia Tech, and Rice University, Falcon presented S2FA (Spark-to-FPGA-Accelerator) at DAC ‘18. S2FA, an automation framework that leverages the Falcon Merlin Compiler, was able to achieve close to 50X performance improvement for several Machine Learning applications.

Download the Full Paper to Learn More
Title
TGPA: Tile-Grained Pipeline Architecture for Low Latency CNN Inference
International Conference on Computer Aided Design November 5, 2018
S2FA: An Accelerator Automation Framework for Heterogeneous Computing in Datacenters
Design Automation Conference (DAC) June 24, 2018
Automated Accelerator Generation and Optimization with Composable, Parallel and Pipeline Architecture
Design Automation Conference (DAC) June 24, 2018
Automated Systolic Array Architecture Synthesis for High Throughput CNN Inference on FPGAs
Design Automation Conference (DAC) June 18, 2017
Programming and Runtime Support to Blaze FPGA Accelerator Deployment at Datacenter Scale
ACM Symposium on Cloud Computing October 5, 2016
Software Infrastructure for Enabling FPGA-Based Accelerations in Data Centers
International Symposium on Low Power Electronics and Design August 8, 2016
Energy-Efficient CNN Implementation on a Deeply Pipelined FPGA Cluster
International Symposium on Low Power Electronics and Design August 8, 2016
Heterogeneous Datacenters: Options and Opportunities
Design Automation Conference (DAC) June 5, 2016
A Novel High-Throughput Acceleration Engine for Read Alignment
International Symposium on Field-Programmable Custom Computing Machines July 16, 2015

Video

Deploying Accelerators At Datacenter Scale Using Spark

Academic Partnerships

Are you part of an academic institution that could benefit from our acceleration platform? Get in touch.

Thank you! The information has been submitted successfully.