Network-on-chip interconnect helps enable deep neural networks for machine vision and learning
CAMPBELL, Calif. – Arteris Inc., the inventor and leading supplier of silicon-proven commercial Network-on-Chip (NoC) interconnect IPsolutions, today announced that Movidius has licensed Arteris® FlexNoC IP for use in its machine learning vision processing units (VPU).
Movidius pioneered the Myriad family of VPUs, with the ability to execute concurrent, mission-critical vision-specific workloads in real time, at high performance and using ultra-low power. Movidius is unique because it has developed its own processing architecture designed specifically for the large-scale numerical workloads of image and vision processing. This highly parallel programming methodology and architecture achieves a "sweet spot" in terms of ultra-low power dissipation for the demanding workloads of augmented and virtual reality, robotics, drones, and smart security.
Movidius selected Arteris FlexNoC interconnect IP for the following reasons:
- Ultra-low power consumption, including the ability to safely disconnect power not only to individual IP blocks, but also to parts of the on-chip fabric and clock tree.
- Advanced simulation and architectural exploration features to obtain the highest bandwidth and lowest latency designs.
- Ability to easily create scalable, hierarchal designs using FlexNoC’s composition features.
“Ultra-low power consumption is one of our key requirements, and Arteris FlexNoC interconnect IP is the best in helping us manage interconnect power, and by extension SoC-wide power,” said Sean Mitchell, co-founder and COO of Movidius. “Combined with the excellent simulation, design exploration, and debugging features, Arteris FlexNoC was the clear choice from both the power-performance-area and time-to-market perspectives.”
“We are proud to contribute to Movidius’s success in visual processing and machine learning, helping expand the capabilities of their highly parallelized SoC architecture,” said K. Charles Janac, President and CEO of Arteris. “Movidius’ use of Arteris FlexNoC for bandwidth-intensive and latency-sensitive SoCs with ultra-low power requirements is proof of Arteris FlexNoC’s scalability and unrivaled power management capabilities.”
Movidius is the leader in high performance, ultra-low power computer vision technology for connected devices. By marrying sophisticated software algorithms to a powerful, purpose-built Vision Processing Unit (VPU), Movidius brings new levels of visual intelligence to smart devices. Movidius’ dedicated machine vision solutions enable a new wave of intelligent and contextually aware devices including drones and AR/VR devices. With offices in Silicon Valley, Ireland, Romania and China, Movidius is a venture-backed company with investors including Atlantic Bridge Capital, Draper Esprit, Robert Bosch Venture Capital, and Summit Bridge Capital.
Arteris, Inc. provides Network-on-Chip interconnect IP and tools to accelerate System-on-Chip semiconductor (SoC) assembly for a wide range of applications. Rapid semiconductor designer adoption by customers such as Samsung, Huawei / Hisilicon, Mobileye, Altera, and Texas Instruments has resulted in Arteris being the only semiconductor IP company to be ranked in the Inc. 500 and Deloitte Technology Fast 500 lists in 2012 and 2013. Customer results obtained by using the Arteris product line include lower power, higher performance, more efficient design reuse and faster SoC development, leading to lower development and production costs. More information can be found at www.arteris.com.
+1 408 470 7300