Kudan ports AI vision algorithm to Tensilica processor core

Kudan ports AI vision algorithm to Tensilica processor core

Business news |
By Nick Flaherty

Cadence Design Systems has signed up Kudan and to its Tensilica software partner ecosystem, bringing simultaneous localization and mapping (SLAM) and AI image signal processor (ISP) to its processor cores.

UK-based Kudan develops visual odometry software and was an early implementer of SLAM algorithms, using cameras to identify position and orientation. While fairly complex, SLAM is implemented in many household products, including robot vacuum cleaners and AR/VR devices, and in more complex systems, such as self-driving vehicles and self-navigating drones.

Kudan’s proprietary SLAM implementation on the Tensilica Vision Q7 DSP yielded a 10X performance improvement for the front-end feature extraction stages, translating to a nearly 15% speedup of the overall SLAM pipeline compared to CPU-based implementations.

“To best address our customers’ design challenges, it’s important to approach solutions holistically. From a hardware standpoint, ongoing innovations in Tensilica IP and architecture are critical for continued momentum with leading smartphone manufacturers and providers of IoT systems and next-generation connected vehicles,” said Amol Borkar, director of product management, marketing and business development for Tensilica Vision and AI DSPs at Cadence.

“A robust software ecosystem is also essential, and partnering with industry leaders like Kudan and helps us bring cutting-edge SLAM and AI ISP solutions to run efficiently on Tensilica IP-based devices with the best power-performance envelope yet.”

“We’re very excited about our partnership with Cadence and the opportunity to work with the Tensilica platform to accelerate Kudan’s SLAM pipeline,” said Juan Wee, CEO at Kudan USA. “Cadence’s Tensilica Vision DSPs provide specialized instructions that optimize various stages of the SLAM algorithm, delivering significant gains with power savings to the end customer. We look forward to improving the accessibility and adoption of our SLAM solution together.”

The Tensilica software ecosystem includes more than 50 partners developing solutions for automotive, smartphone apps, IoT, software services and other segments.

In most systems with image or video functionality, the set of ISP functions is well established and typically provided as a hardware block, resulting in high-performance throughput but no flexibility or adaptability to current conditions.’s novel approach uses AI to replace traditional hardwired ISP functions, enabling high-quality videos to be produced in real-time, even in the most challenging lighting conditions.’s efficient AI-ISP enables customers to implement a camera pipeline with resolutions greater than full HD while operating at over 30fps on the Tensilica NNA110 accelerator.

“At, we have developed a method of using AI to dramatically improve image quality in real time, particularly in the most challenging lighting conditions,” said Oren Debbi, CEO of “For this technology to reach its true potential, there is a need for fast and efficient neural network computations. Joining Cadence’s Tensilica ecosystem will help ensure that our customers have a very competitive solution that runs on some of the most efficient vision and AI platforms out of the box.”

If you enjoyed this article, you will like the following ones: don't miss them by subscribing to :    eeNews on Google News


Linked Articles