Learn how to implement Cosine Similarity for TigerGraph accelerated graph analytics running on HPE DL385 ProLiant servers with certified Alveo U50 accelerator cards, in this reference design document.
Join TigerGraph in a live webinar to learn how Xilinx turbocharges Customer Journey/360 performance for healthcare organizations. Register now.
Gain real-time insights from your connected data with TigerGraph. Try out our free Customer Journey/360 CosineSim tool kit on Github to get started.
When it comes to patient care, providing care path recommendations in real time is essential to enabling better patient outcomes.
Imagine if you could provide a better patient experience, with insights delivered in the rhythm of a natural conversation and make your call center more efficient at the same time.
TigerGraph, the world’s fastest and most scalable graph analytics platform, enables you to connect, analyze, and learn from your siloed patient data.
Now with Xilinx acceleration, TigerGraph delivers critical care path recommendations in seconds, not minutes.
Before TigerGraph, the only way to identify patient similarity was to log on to 15 different systems, and spend hours, if not days, determining a personalized care path.
Now, with the click of a button, you can create a real-time 360° view of your members’ journey – improving their quality of care.
TigerGraph enables you to drill down deep into member data without sacrificing performance. TigerGraph easily finds similarities up to 20 hops deep, even across disparate systems, and Xilinx acceleration delivers these results in seconds, allowing your representatives to provide your members with the comfort of instant, accurate answers.
The Xilinx Cosine Similarity algorithm for TigerGraph dynamically personalizes clinical outcome predictions by identifying and analyzing similar past patients, rather than accessing all available patient records.
With Xilinx, loading and analyzing terabytes of data with ten or more hops, is completed in seconds.
Watch this demo video to learn more.
Turbocharging your TigerGraph deployment with Xilinx is simple. Just deploy off-the-shelf Xilinx Alveo accelerator cards, fully qualified for this application on HPE ProLiant DL385, and realize a 300x performance gain on TigerGraph at 95% lower TCO compared to running on a CPU.
Along with the HPE reference design for ready-to-go pre-qualified servers, all the tools, user guides and installation tips are available on GitHub.
Customer Journey/360 Cosine Similarity tool kit on GitHub:
- Customized User Defined Function (UDF) plug-ins to GSQL
- C/C++, Python and Jupyter Notebook
- Xilinx Vitis™ development flow
TigerGraph graph analytics is available in a HPE DL385 Gen 10+ server, populated with up to 8 x Alveo U50 accelerator cards.
Join our subscriber list for the latest updates on Data Analytics from Xilinx and our partners.