Innovations in Trading and TechnologySponsored by NYSE Technologies
What It Takes to Build a Successful In-database Analytics Platform
28 March 2011:
Rather than pulling data to where the analytics are, why not push the analytics to where the data are? Dan Watkins, vice president, financial institutions at Fuzzy Logix, explains the time and cost savings and better results of doing so.
For more videos in the Innovations in Trading and Technology Spotlight Series click here.
Related Videos
Rate This VideoYour Rating
Login to rate.
|
Community Rating
|
Comments | Post a Comment
2 Comments to "What It Takes to Build a Successful In-database Analytics Platform":
David N. W. Sawyer
30 March 2011
Dan, Greg, I think I now understand the Fuzzy value prop of faster analytics colocated on an FPGA powered data appliance. But can you elaborate on the AI side of the equation - are we now seeing algo traders gaining an edge by applying predictive analytics to series of data that were previously too big to work with in a time effective fashion ?
dwatkins
05 April 2011
David, thanks for stopping in. FPGA's are involved some of the time with certain players, but thats not a concern of Fuzzy Logix necessary. Our point is, most analytics are done outside of the database environment making it a smaller and more timely task. Run analytic functions where the data is stored eliminating the need to remove information can give 100x's faster results and sometimes as high a 30:1 return on investment when doing so.
Comments (8)