With a fundamentally different architecture than legacy products, KNUPATH can operate alone or be integrated with other devices, and it is available now to both end customers and technology vendors seeking to create data center neural computing capabilities to support advancements in machine learning, IoT and signal processing.
"Many of today's CPUs, GPUs and FPGAs force system designers to either create workarounds with last-generation chip sets or reduce their requirements for advanced-compute projects," said Dan Goldin, Founder and CEO of KnuEdge. "After ten years of stealth development and rigorous testing, LambdaFabric enters the market as mature technology which enables system designers to meet the most demanding requirements now, and also helps them rethink what is possible with neural computing in the future."
As evidenced by recent announcements such as Google's Tensor Processing Unit, there is increasing industry interest in targeted processor acceleration for machine learning and other growing workloads. KNUPATH LambdaFabric is inherently designed to scale up to 512,000 devices and beyond, and offers rack-to-rack latency of only 400 nanoseconds, approaching half the latency of existing high speed interconnects. KNUPATH processors are low-wattage, 256-core chips with 16 bi-directional I/O paths delivering 320 Gbs-per processor. In a watt-to-watt comparison with today's processing technology, the first generation of LambdaFabric processors is demonstrating a performance advantage of 2x to 6x in targeted environments.
Based on Neurobiology
KNUPATH delivers entirely new technology based on neurobiological principles, and it will reset the standard for targeted signal processing and machine learning in data centers and IoT devices. KNUPATH was created when KnuEdge encountered a mission-critical roadblock while developing its neural voice recognition and authentication technology. The KnuEdge team realized it could not achieve the performance it would need with traditional CPUs and GPUs, and created a new team to build an application-specific integrated circuit (ASIC). As LambdaFabric was developed, it became clear that this technology would be broadly applicable for data centers around the world.
"As organizations develop more machine learning, IoT, signal processing and other compute-intensive applications, the software will only be as good as the hardware we put it on," said Patrick Patla, SVP and GM of KNUPATH. "LambdaFabric removes many of the common hardware barriers system designers have faced in the past several years, such as super-scalability and inter-platform latency. The market is just beginning to understand what is possible with this technology, and we're looking forward to working with a broader group of the world's system design teams to discover new opportunities in data centers and with edge devices."
KNUPATH will deliver production platforms and systems in the second half of 2016. For demos of KNUPATH or for a more in-depth LambdaFabric processor overview, visit
www.knupath.com.
Related News: World-renowned Aerospace and Tech Visionary Launches KnuEdge Today
Before today's official launch of KnuEdge, only a couple hundred people had heard about the company. By design, its small, geographically dispersed teams spent a decade in stealth mode developing two products to maturity: KNUPATH and KnuVerse, a next-generation voice recognition and authentication platform, which was also
announced today.
The company has more than $100 million in private investment and is founded by Dan Goldin, the world-renowned innovator who spent the past five decades innovating and delivering aerospace and technology products years ahead of the curve. He also led the National Aeronautics and Space Administration (NASA) throughout its renaissance in the 1990s. Goldin's tenure at NASA reads like a Harvard Business Review case study in successful execution. For instance, as NASA chief, he guided the redesign and delivered the International Space Station, initiated bold robotic exploration of Mars and formed the Astrobiology Institute to better understand the origin, evolution and destiny of life in the universe. In addition, he put a record number of people into space without incident, all while reducing the agency's planned budget by 33%.
"At KnuEdge, we are not in business to create incremental technology improvements on what already exists. Our mission is fundamental transformation," said Goldin. "We were swinging for the fences from the very beginning, with intent to create technologies that will in essence alter how humans interact with machines, and enable next-generation computing capabilities ranging from machine learning to artificial intelligence."
