ARMONK, N.Y., Jan. 14, 2014 /PRNewswire/ -- IBM (NYSE: IBM) today announced that its inventors received a record-setting 6,809 patents in 2013 – the 21 st consecutive year the company topped the annual list of U.S. patent leaders.
IBM's 2013 patent results represent a diverse range of inventions poised to enable significant innovations that will position the company to compete and lead in strategic areas–such as IBM's Watson, cloud computing, Big Data and analytics. These inventions will also advance the new era of cognitive systems where machines will learn, reason and interact with people in more natural ways.
"We take pride in being recognized as the U.S. patent leader, but patents are only one gauge of innovation. Equally significant is the impact that our patented inventions have when they are used to enable solutions that help clients and societies solve problems," said Bernie Meyerson, IBM Fellow and VP of Innovation. "Furthermore, the broad range of inventions that these patents represent underscores the need for a patent system that equally and fairly promotes and supports innovation across all technical fields."IBM's 2013 patent total exceeded the combined totals of Amazon, Google, EMC, HP, Intel, Oracle/SUN and Symantec. The company's record 2013 patent count was made possible by more than 8,000 IBM inventors residing in 47 different U.S. states and 41 countries. The Top Ten list of 2013 U.S. patent recipients* includes:
- IBM 6,809
- Samsung 4,676
- Canon 3,825
- Sony 3,098
- Microsoft 2,660
- Panasonic 2,601
- Toshiba 2,416
- Hon Hai 2,279
- Qualcomm 2,103
- LG Electronics 1,947
- U.S. Patent # 8,510,296 Lexical answer type confidence estimation and application – This patented invention enables IBM Watson to more accurately assess questions posed in natural language and determine confidence in the accuracy of potential answers. To accelerate the impact of patented Watson inventions, IBM has formed the Watson Group, a new business unit that will tap the company's software, services, research, industry experts and sellers to advance development and delivery of a new class of Watson-enabled cognitive computing apps and technologies to the marketplace.
- U.S. Patent #8,515,885: Neuromorphic and synaptronic spiking neural network with synaptic weights learned using simulation – This patented invention describes breakthrough brain-inspired computers that lay the foundation for a new generation of cognitive systems via hardware and software co-design. As part of a DARPA-funded Systems of Neuromorphic Adaptive Plastic Scalable Electronics (SyNAPSE) project, IBM is developing a chip architecture that aspires to emulate the human brain's cognitive capabilities while rivaling its energy and volume efficiency.
- U.S. Patent #8,422,686: Automated validation and execution of cryptographic key and certificate deployment and distribution – This patented invention automates the lifecycle of cryptographic keys used to encrypt and secure data – from creation and deployment to deletion and can also enhance security for cloud computing applications.
- U.S. Patent #8,352,953: Dynamically Provisioning Virtual Machines – This invention solves the " noisy neighbor" problem that reduces online system availability and constrains cloud computing network bandwidth when websites, such as online retailers or auction sites encounter unexpected dramatic spikes in demand.
- U.S. Patent #8,387,065: Speculative popcount data creation - This patent describes an approach for Big Data and analytics computing where a small region or population of analyzed data -- known as a popcount -- is counted, sorted, and speculatively analyzed in real time for trends or outliers. The idea is based on a counter intuitive premise: that Big Data analysis is small. The method improves data analysis performance, reduces processor resources needed to analyze the data and is based upon modern graph theories. Built into IBM's POWER processors -- the chip within IBM's Watson and IBM's Power Systems servers -- speculative popcount is advancing cryptanalysis, real-time error correction for streaming data and the cognitive computing era.
- U.S. Patent #8,423,339: Visual analysis of a protein folding process – This patented invention describes a method for discovering and viewing common patterns in protein folding simulation, which aids in understanding the protein folding process and can lead to significant advances in computer based drug discovery, among other applications.
- U.S. Patent #8,572,274: Estimating load shed data in streaming database application – This patented invention describes a technique that addresses the challenge of analyzing real-time streaming Big Data traffic jams by shedding or reducing the data pool without compromising accuracy of the insights.
- U.S. Patent #8,402,041: Analytics of historical conversations in relation to present communication – This patented invention uses Big Data analytics to establish relationships between past and present electronic conversations (social or business) that have similar attributes and provide relevant results in real-time.