Technology Trends in 2013-2014
At the Technology Trends seminar, with HCMC University of Polytechnics' lecturers, KMS Technology's CTO delivered a topic of Big Data, Cloud Computing, Mobile, Social Media and In-memory Computing.
1. 1TECHNOLOGY TRENDS FOR 2013Kaushal Amin, Chief Technology OfficerKMS Technology Atlanta, GA, USA 2. ABOUT KMS2Founded in January 2009 with offices inAtlanta, Dublin, Calif., and Ho Chi Minh City, Vietnam, KMSTechnology is a US Offshore Product Development (OPD)company.We have a 400+ global workforce that provides a variety ofcommercial grade web and software development services tosoftware product and technology-based companies. 3. ABOUT SPEAKER KAUSHAL AMIN32011-NowKMS2006-11LexisNexis2001-06Startups1999-02Intel1993-99McKesson1989-93IBM1985-88Engineering Bachelors inComputer Engineeringfrom University ofMichigan Developed OS CrossAssembler in C forMC6809 Developed WindowsNT based optical filesystem for dealingwith large data files Healthcare MedicalRecords & Imaging Wireless mobile fieldservice software onWindows CE and J2ME Developed PriceOptimization softwarefor retail and hotelindustry Provide technicalleadership andmentoring to KMS USand Vietnam staff Provide C leveltechnology consultingto KMS clients Part of OS/2 Kernelteam Atlanta Police MobilePlatform (Motorola) Delta Flight Planning& Fueling Systems inUnix Intels multimediashowcase website in16 languages and 40+countries One of the early N-tierarchitected WindowsCOM+ web system Online BIG DATAsystem of US criminalrecords, education, and employment historyon employees LexisNexis s NoSQLdistributed database 4. WHY SHOULD YOU BE HERE Learn about MAJOR software technology trends affecting ITindustry and businesses Necessary in order to anticipate and respond to ongoingtechnology-driven disruptions Step up. Provoke and harvest disruption. Dont get caught unawareor unprepared.4 5. INDUSTRY EXPERTS 2013 LIST5 6. #1 MOBILE APPS6 Mobile devices overtaking PCs as the most commonweb access device worldwide by end of 2013 More market shift towards complex businessapplications instead of small niche consumer apps Similar to PC evolution of desktop productivity apps tonetwork enabled enterprise solutions Apple iOS and Google Android will continue todominate market share for next 2 years Native Apps will continue to be preferred developmentplatform, however, HTML5/Hybrid will start gainingground 7. MOBILE APPS STATS7Mobile App Market Stats: The number of smartphones will exceed 1.82 billion unitsworldwide in 2013 Android is expected to claim 63.8% market share by 2016 iOS monthly revenues are 4x those of Google Play Apple has paid developers $5 billion in app sales There are now more than 400 million accounts withregistered credit cards in the App Store Google Play Has 700,000 Apps, Tying Apples App Store 8. #2 - BIG DATA8 Automatically generated by a machine(e.g. Sensor embedded in an engine) Typically an entirely new source of data(e.g. Use of the internet) Not designed to be friendly(e.g. Text streams) May not have much valuesNeed to focus on the important part 9. BIG DATA - NOSQL9 Next Generation Databases mostly addressingsome of the points: being non-relational, distributed, open-source andhorizontal scalable. Key factor over SQL databases is its ability to storeand retrieve data across multiple commodity servernodes in parallel The original intention has been modern web-scaledatabases. The mass movement began early 2009 and isgrowing rapidly. However, core technology datesback to 1990s. 10. BIG DATA TECHNOLOGIES10 MapReduce Technique for indexing andsearching large data volumes Google Invention, Hadoop Column Store Each storage block contains datafrom only one column HBase, Cassandra Document Store Stores documents made up oftagged elements MongoDB, CouchDB Key-Value Store Hash table of keys Berkley-DB, Voldemort 11. BIG DATA STATS11 Google processes 100 PB/day; 3 million servers Facebook has 300 PB + 500 TB/day; 35% ofworlds photos YouTube 1000 PB video storage; 4 billionviews/day Twitter processes124 billion tweets/year SMS messages 6.1T per year US Cell Calls 2.2T minutes per year 12. #3 - CLOUD COMPUTING12 Shift from Should we use to how can we usecloud within corporate IT Personal Cloud to replace PCs for personal contentstorage allowing access across multiple devices Cloud-based disaster-recovery as-a-service De-duplicating and Encryption of data before it issent to a cloud storage service will be an integralcomponent 13. CLOUD COMPUTING13 Start addressing the real drawbacks of cloudcomputing - the challenges of scale, complexity andchange management - rather than fixating on itssupposed drawbacks such as security, compliance andSLAs SaaS applications will continue to be developed usingCloud Computing (private or public) 14. #4 - IN-MEMORY COMPUTING14Enabling users to develop applications that run advancedqueries or perform complex transactions, on very largedatasets, at least one order of magnitude faster and ina more scalable way than when using conventionalarchitectures- Gartner definitionExamples: Fraud Detection Price Optimization Demand Forecast Flight Control Fueling, Maintenance, & Scheduling Simulation (What-If Analysis) 15. IN-MEMORY COMPUTING15Why Now? 64-bit processors allowing access to 16 exabytes ofmemory (32-bit limited it to 4GB) Memory chips getting faster, more capacity, andcheaper due to Moores law New off-the-shelf commodity servers are capable of1TB RAM capacity big enough for many largedatabases to remain in memory In-Memory RDBMS from Oracle, Microsoft, and othersallowing traditional SQL based applications to benefitimmediately by placing data in memory New development tools making it easier for developersto build applications running across multiple bladeservers e.g. 1000 servers 4 cores per server with 512 GB RAM 16. IN-MEMORY COMPUTING16 In-Memory Computing can squeeze batch processesnormally lasting hours into minutes or seconds. These processes are provided in the form of real-timeor near real-time services and delivered to users in theform of cloud services. Numerous vendors will deliver in-memory solutionsover the next two years, driving this approach intomainstream use. 17. #5 - ACTIONABLE ANALYTICS17 To make analytics more actionable and pervasivelydeployed, BI and analytics professionals must makeanalytics more invisible and transparent to their users Embedded analytic applications at the point ofdecision or action Real-time operational intelligence systems that makesupervisors and operations staff more effective Provides simulation, prediction, optimization and otheranalytics, to empower even more decision flexibility atthe time and place of every business process action Enabled by Big Data and In-Memory Computingtechnologies 18. ACTIONABLE ANALYTICS18Tools: Google Analytics Teradata Greenplum Woopra Juice Analytics Jaspersoft KISSmetricsExamples: Improving Quality of Healthcare Leveraging CRM data at the point of sell (Amazon) Gaining Operational Efficiency Field Service Order Processing 19. #6 SOCIAL MEDIA19 Social Media trend continues to grow and morebusiness applications will leverage social mediathrough integrations The three most trusted forms of advertising are: Recommendations from people I know - 90% Consumer opinions posted online - 70% Branded websites - 70% Mobile in the middle and primary device for use ofsocial media Google+ Is a Must - Google+ integration now extendsto many Google properties, such asYouTube, Gmail, Blogger, and Search 20. 20MOST USED SM TOOLS 21. NEXT STEPS Step Up. Expand your knowledge about what interests you themost pick 3 areas Provoke and harvest disruption. Dont get caught unaware orunprepared Look for Game Changer opportunities within your projects throughuse of technologies Keep in Mind - Your projects may not adopt or use all of thetechnologies21 22. 2013 KMS TechnologyQ&A