Quadric, a leading processor technology intellectual property (IP) licensor, today announced that Steve Roddy, a former vice president with Arm’s machine learning group, has joined the company’s executive team as chief marketing officer (CMO). Coinciding with this executive appointment, Quadric also announced that it is bringing its edge AI processor architecture to market as a licensable intellectual property (IP) product.
Roddy brings more than 30 years of marketing and product management expertise across the machine learning (ML), neural network processor (NPU), microprocessor, digital signal processor (DSP) and semiconductor IP industries. He joined ARM in 2018 as vice president in the Machine Learning Group. Prior to Arm, Roddy served in VP roles in IP licensing businesses at Tensilica (Cadence Design Systems) and Amphion Semiconductor, as well as earlier product management roles at Synopsys, LSI Logic and AMCC. Roddy holds a Bachelor of Science degree in electrical engineering from the University of California, Berkeley, and an MBA from UCLA.
“Steve’s exceptional IP product knowledge and licensing go-to-market vision will strengthen our continued momentum in becoming one of the world’s leading processor architectures for on-device ML inference,” said Veerbhan Kheterpal, co-founder and CEO of Quadric. “As Quadric launches the availability of its processor architecture as a licensable IP product and expands its reach across global markets, Steve’s expertise and leadership will be vital to take the company to the next level.”
“Joining Quadric presents a tremendous opportunity to apply my NPU and DSP expertise to what lies ahead for both Quadric and the overall industry,” said Steve Roddy. “Quadric's architecture—already proven in silicon—is the only new processor that can efficiently run both DSP code and machine learning graph inference on the same core. The massive increase in software developer productivity gained by having only one toolchain and one core to program, manage and debug is market changing. I am excited to join the highly accomplished and committed Quadric team at a time when on-device AI computing is catapulting the processor market forward.”
GPNPU – A New Category of Processor IP
The Quadric architecture is the industry's first general-purpose neural processing unit (GPNPU). Quadric's architecture delivers high ML inference performance but—unlike other neural network accelerators that support a limited number of ML graph operators—the Quadric solution also has a general-purpose control and signal processing capability, blending the best attributes of NPU accelerators with DSPs. Quadric GPNPUs can run both neural net graphs and C++ code for signal pre-processing and post-processing.
“In embedded applications, neural networks are usually found alongside classical computer vision, image processing and signal processing algorithms,” said Jeff Bier, founder of the Edge AI and Vision Alliance. “Quadric’s approach of addressing both neural network and other types of algorithms with one processor is quite interesting. I look forward to hearing more about the GPNPU from Quadric in their presentation later this week at the Embedded Vision Summit.”
Quadric's unique ability to handle both neural backbones and classical dynamic, data-parallel algorithms in a unified architecture will help create AI for everyone, everywhere. Quadric’s unified architecture is flexible enough to accelerate the entire data pipeline without the need for complex multicore subsystems and multiple toolchains.
Future Proof SoC Designs
The Quadric GPNPU offers an unmatched degree of flexibility for semiconductor companies selecting processor IP for their new system-on-chip (SoC) developments. Silicon vendors today have to select and license processor engines that will ship in devices sold years from now and be programmed by developers for years after. As machine learning algorithms evolve, Quadric's GPNPU delivers future-proof programmability to run new neural net operators or new C++ code to match the future needs of both data scientists and embedded developers.
Quadric is building a unified hardware and software architecture optimized for on-device AI computing. Quadric offers an integrated full-stack software and hardware platform with optimized algorithms and next-generation GPNPU processor technology that delivers unmatched capabilities and efficiency across applications for a variety of devices. With AI becoming a business necessity in the global economy, customers need complete AI solutions deployed at scale. Quadric's unique ability to handle both neural backbones and classical dynamic data-parallel algorithms in a unified architecture is helping to create AI for everyone, everywhere. Learn more at quadric.io.