gpu-accelerated analytics: case studies and use cases

11
GPU-Accelerated Analytics Enable New Enterprise Solutions 11/30/16 Case Studies and Use Cases

Upload: kinetica

Post on 15-Jan-2017

45 views

Category:

Software


4 download

TRANSCRIPT

Page 1: GPU-Accelerated Analytics: Case Studies and Use Cases

GPU-Accelerated Analytics Enable New Enterprise Solutions

11/30/16

Case Studies and Use Cases

Page 2: GPU-Accelerated Analytics: Case Studies and Use Cases

Enabling New Enterprise Solutions

2

Retail: Customer 360/customer sentiment, supply chain optimizationCorrelating data from point of sales (POS) systems, social media streams, weather forecasts, and even wearable devices. Better able to track inventory in real time, enabling efficient replenishment and avoiding out-of-stock situations.

Powering High Performance “Analytics as a Service” SolutionDelivering customer-focused services by leveraging all available transactional data. Currently no ability for business user to do customized analytics; IT has to. Query response times taking 10s of minutes, some over 2 hours, thus limiting ability to analyze and use data.

Fin servicesLarge scale risk aggregations and billion+ row joins in sub-second time (5TB+ tables choke on RDBMS joins and Hadoop is too slow).

Manufacturing IoTLive streaming analytics on component functionality to ensure safety (avoid failures) and validate warranty claims .

Ridesharing/Connected CarsView all passengers and drivers to monitor behavioral analytics. Sensor enabled predictive maintenance on connected cars.

Presenter
Presentation Notes
Use cases in play include: Real-time sentiment analysis Real-time anomaly detection to prevent fraud Reallocate resources on the fly based upon personnel, environmental and seasonal data Track terrorist and other national security threats in real time Optimize energy generation and uptime based on fluctuating usage patterns and unpredictable natural disasters Improve customer engagements by correlating data from point of sales (POS) systems, social media streams, weather forecasts, and other data. Track inventory in real-time, enabling efficient replenishment and avoiding out-of-stock situations Connected home, connected car, connected cities More rapid query response time and data discovery capabilities for end users Faster iterations to risk modeling and modeling optimization
Page 3: GPU-Accelerated Analytics: Case Studies and Use Cases

US Army INSCOM

US Army’s in-memory computational engine for any data with a geospatial or temporal attribute for a major joint cloud initiative within the Intelligence Community (IC ITE).

Intel analysts are able to conduct near real-time analytics and fuse SIGINT, ISR, and GEOINT streaming big data feeds and visualize in a web browser.

First time in history military analysts are able to query and visualize billions to trillions of near real-time objects in a production environment.

Major executive military and congressional visibility.

Oracle Spatial (92 Minutes)

42x Lower Space28x Lower Cost38x Lower Power Cost

U.S Army INSCOM Shift from Oracle to Kinetica

Kinetica(20ms)

1 Kinetica server vs 42 servers with Oracle 10gR2 (2011)

CASE STUDY

Presenter
Presentation Notes
Game changer for military; first time in history they can fuse UAV live track data with SIGINT data to target on the fly with high accuracy 1 server with GPUdb can replace 42 server with Oracle 10gR2. 42 servers doing a geospatial query took 92 minutes; same data, same query 1 GPUdb server took less than 1 second
Page 4: GPU-Accelerated Analytics: Case Studies and Use Cases

Cybersecurity

4

CASE STUDY

Real-time anomaly detectionFusing multiple data feeds.Protecting its financial clients against current and emerging cyber threats.

Presenter
Presentation Notes
US ARMY again mention USPS is the single largest logistic entity in the country, moving more individual items in four hours than the combination of UPS, FedEx, and DHL move all year. Distributed analysis - USPS’ parallel cluster is able to serve up to 15,000 simultaneous sessions, providing the service’s managers and analysts with the capability to instantly analyze their areas of responsibility via dashboards and to query it as if it were a relational database. At scale With 200,000 USPS devices emitting location once every minute, that amounts to more than a quarter billion events captured and analyzed daily tracked on 10 nodes. Ironnet several major banks, and Sis PG&E - Data silos between gas, electric, distribution/transportation, fiber vs land-based.We can consolidate feeds, fuse analytics, proactive on utility data, prevent crisis, full real time visibility on workforce, meters, grids. �
Page 5: GPU-Accelerated Analytics: Case Studies and Use Cases

Real-time fleet optimization

DISTRIBUTED ANALYSIS

USPS’ parallel cluster is able to serve up to 15,000 simultaneous sessions, providing managers and analysts with real-time dashboards

AT SCALEWith 200,000 USPS devices emitting location once every minute, that amounts to more than a quarter billion events captured and analyzed daily.

USPS–the largest logistic entity in the US–tracks carrier movements in real time.

Reallocates resources on the fly based upon personnel, environmental and seasonal data.

CASE STUDY

In production since 2014 with 5 9’s uptime

Presenter
Presentation Notes
Been running since Nov 2014 at 5 9’s
Page 6: GPU-Accelerated Analytics: Case Studies and Use Cases

Real-time fleet optimization

Analyzing Breadcrumb Data to Improve Services 200,000 USPS devices emit location once every minute, amounting to more than a quarter billion events daily.

By analyzing this data, the USPS is able to 1) understand where spending would

achieve the best results2) make faster and more efficient strategic

decisions 3) provide customers with a reliable service4) reduce costs by streamlining deliveries

CASE STUDY

Processing Geospatial Data for Real-Time Decision-makingDispatchers can graphically view employee territory assignments (routes, delivery points, and collection points).

The USPS makes better use of routes by: 1) finding inefficiencies such as overlapping

coverage of assigned areas, uncovered areas, and distribution bottlenecks

2) improving contingency planning if a carrier was unable to deliver to assigned routes

3) using their workforce more efficiency by aggregating point-to-point carrier performance data

Wins IDC HPC Innovation Excellence Award, 2016

Presenter
Presentation Notes
Been running since Nov 2014 at 5 9’s
Page 7: GPU-Accelerated Analytics: Case Studies and Use Cases

Redesign Retail without Redesigning your Data

Multidimensional query on “product category x” sales on map.

Drill down to any geographic area for enhanced detail.

CASE STUDY

Page 8: GPU-Accelerated Analytics: Case Studies and Use Cases

Smart Grid Visualization and Optimization

8

Bringing real-time visibility and analytics to legacy systemGeospatially visualizing smart grid while ingesting real-time vector data streaming from smart meters, workforce and grid.

Decreasing current SLA for geospatial analytics. Leveraging open standards. Consolidating feeds and layering multiple datasets on the same visualization, fusing analytics.

Striving for optimized energy generation and uptime based on fluctuating usage patterns and unpredictable natural disasters – and crisis prevention.

CASE STUDY

Presenter
Presentation Notes
perform geo-spatial queries (freehand-polygon for example) that current system cannot do, have on the fly rendering of heat maps, flows, etc… with real time drill down to an individual component (meter, transformer, pipe, etc…) PG&E - Data silos between gas, electric, distribution/transportation, fiber vs land-based.We visualize billions of points, �
Page 9: GPU-Accelerated Analytics: Case Studies and Use Cases

Retail Solutions

Customer 360/Customer SentimentOut think the status quo. Correlate data from point of sales (POS) systems, social media streams, weather forecasts, and even wearable devices. Convert billions of impressions to millions of dollars.

Be able to better understand your customers and your business by querying massive datasets in seconds vs. hours to help you become a real-time retail enterprise.

Supply Chain OptimizationMeet demand spikes in real time. Better track inventory in real time from a variety of sources, enabling efficient replenishment and avoiding out-of-stock situations.

Presenter
Presentation Notes
Use cases in play include: Real-time sentiment analysis Real-time anomaly detection to prevent fraud Reallocate resources on the fly based upon personnel, environmental and seasonal data Track terrorist and other national security threats in real time Optimize energy generation and uptime based on fluctuating usage patterns and unpredictable natural disasters Improve customer engagements by correlating data from point of sales (POS) systems, social media streams, weather forecasts, and other data. Track inventory in real-time, enabling efficient replenishment and avoiding out-of-stock situations Connected home, connected car, connected cities More rapid query response time and data discovery capabilities for end users Faster iterations to risk modeling and modeling optimization
Page 10: GPU-Accelerated Analytics: Case Studies and Use Cases

Automotive Solutions

RidesharingView all passengers and drivers to monitor behavioral analysis. Watch for fast acceleration, sudden braking, too many U turns, etc. to avoid risk/lawsuits of faulty drivers.

Connected carsSystems monitoring, live streaming ingest and analytics for sensor enabled predictive maintenance. Automated emergency management, smart driving assistance, and real-time fleet managment

Presenter
Presentation Notes
Use cases in play include: Real-time sentiment analysis Real-time anomaly detection to prevent fraud Reallocate resources on the fly based upon personnel, environmental and seasonal data Track terrorist and other national security threats in real time Optimize energy generation and uptime based on fluctuating usage patterns and unpredictable natural disasters Improve customer engagements by correlating data from point of sales (POS) systems, social media streams, weather forecasts, and other data. Track inventory in real-time, enabling efficient replenishment and avoiding out-of-stock situations Connected home, connected car, connected cities More rapid query response time and data discovery capabilities for end users Faster iterations to risk modeling and modeling optimization
Page 11: GPU-Accelerated Analytics: Case Studies and Use Cases

Manufacturing IoT Solutions

Manufacturing IoT• Live streaming analytics on component functionality.• Ensure safety and avoid failures• Track and monitor inventory, materials and operations • Defect detection• Track quality, returns, warranty claims

Presenter
Presentation Notes
Use cases in play include: Real-time sentiment analysis Real-time anomaly detection to prevent fraud Reallocate resources on the fly based upon personnel, environmental and seasonal data Track terrorist and other national security threats in real time Optimize energy generation and uptime based on fluctuating usage patterns and unpredictable natural disasters Improve customer engagements by correlating data from point of sales (POS) systems, social media streams, weather forecasts, and other data. Track inventory in real-time, enabling efficient replenishment and avoiding out-of-stock situations Connected home, connected car, connected cities More rapid query response time and data discovery capabilities for end users Faster iterations to risk modeling and modeling optimization