SANTA FE, N.M.--(BUSINESS WIRE)--Knowm Inc., a start-up pioneering next-generation advanced computing architectures and technology, today formally launched as a company and announced plans to build the world’s first truly adaptive neuromemristive processor. Thermodynamic RAM (kT-RAM) is based on the theory of AHaH computing - a new advanced computing approach inspired by nature and biological nervous systems. The result of more than 13 years of research and development, AHaH Computing eliminates the duality of memory and processing, and provides the architecture for building a new type of ‘soft-hardware’ that self-repairs and self-learns from incoming data streams within the environment. The result is power efficiency gains of up to 10 orders of magnitude over traditional computing architectures. In addition, the company announced the commercial availability of its first products and the Knowm Developer Program.
“If big data has taught us anything it's that there are limitations in performing at scale, particularly with projects requiring the use of machine learning to perform complex tasks in near real-time, such as the detection of security anomalies within petabytes of network traffic,” said Brad Shimmin, service director, Business Technology and Software at Current Analysis. “Knowm’s unique approach to ‘soft-hardware’ with neuromemristive processors presents the industry with some intriguing opportunities to create highly differentiated applications of advanced analytics capable of scaling to meet even the most extreme data requirements.”
The availability of kT-RAM will have the largest impact in fields that require higher computational power such as machine learning and artificial intelligence by massively increasing speed and power performance at the scale of biological systems. In the same way that GPUs greatly accelerated graphics applications, kT-RAM will accelerate machine learning applications, opening up entirely new possibilities on autonomous and mobile platforms and reducing data center operating costs. Others have turned to quantum computing, parallel CPU architectures, GPUs and horizontally-scaled clustering as potential approaches to overcome these limitations. However, those approaches face their own challenges and constraints despite decades of research and billions of dollars invested, making it unlikely they will be ready for mass adoption in the near term, if ever.
“We’ve reached a tipping point in computing as the market is discovering the exciting potential of artificial intelligence applications while concurrently Moore’s Law is being halted in its tracks due to physical size limitations of transistors,” said Tim Molter, CTO, Knowm Inc. “kT-RAM and similar architectures based on the theory of AHaH Computing have the potential to lead the next generation of computing architectures and applications. With kT-RAM, we will see a quantum leap in power, speed and size metrics, leaving competing technologies far behind.”
Knowm believes nature is the highest form of technology and has used that belief to solve a fundamental problem inherent in modern computing architectures when applied to machine learning. Knowm seeks to eliminate the duality of memory and processing with its introduction of AHaH Computing and a new synaptic processor known as kT-RAM, allowing for higher performance computing tasks to be undertaken while maintaining manageable power and infrastructure costs for more advanced applications. AHaH Computing embraces hardware as part of the solution and designs from the bottom-up (hardware) and top-down (algorithms). The result is biologically-efficient machine learning systems that are not only broadly commercially viable, but represent a significant cost and energy savings over current approaches.
AHaH Computing’s roots trace to 2002 when Knowm’s co-founder and CEO, physicist engineer Alex Nugent, discovered mechanisms common to both living and non-living natural systems that support brain-like operations required in the acts of perception, planning and control. Ultimately, Nugent patented what he called the “Knowm synapse,” which helped inspire the creation of the DARPA SyNAPSE and Physical Intelligence programs, which he advised, with direct support later provided by the Air Force and Navy. These physically adaptive mechanisms, used by nature to enable growth and repair, are found in everything from the neurons within the human brain to trees, plants, rivers and lightning.
“Everything eventually reaches its physical limits. We are currently witnessing the end of Moore’s Law scaling in CMOS electronics just as we are finally gaining the computational capacity needed to demonstrate the potential of machine learning,” said Alex Nugent, CEO and co-founder of Knowm. “Pursuing the same digital computing methodologies for machine learning systems keeps us from realizing the full potential of machine learning and artificial intelligence. Through AHaH Computing, Knowm is making hardware soft, uniting memory and processing for synaptic operations, and in the process helping to define the post-Moore’s Law era.”
For more information about Knowm, please visit www.Knowm.org. Like Knowm on Facebook (https://www.facebook.com/knowmorg?ref=hl) and follow Knowm on Twitter (https://twitter.com/knowmorg), Google+ (https://plus.google.com/+KnowmOrg/posts), and LinkedIn (https://www.linkedin.com/company/knowm-inc).
About Knowm Inc.
Knowm believes nature is the highest form of technology and has used that belief to solve a fundamental problem inherent in modern computing architectures when applied to machine learning. The company has taken the best of machine learning and quantum computing to define a new form of computing – AHaH Computing. Knowm is ushering in the next generation of physically adaptive computing hardware with the industry's first general purpose neuromemristive processor specification and development platform, Thermodynamic RAM (kT-RAM), and the availability of discrete memristors, Back End of Line (BEOL) Memristor services, the SENSE and Application Servers, and the Knowm API and Knowm Anomaly. For more information about Knowm, please visit www.knowm.org.