Synaptics launches AI vision and audio module, KatanaConnect

Vision

Edge developer Synaptics has launched its KatanaConnect module, a wireless AI vision and audio device.

Synaptics specialise in human interface hardware and software, with KatanaConnect using AI computer vision and audio sensing to support applications ranging from security to automation and monitoring.

Ananda Roy, senior product manager of low-power edge AI at Synaptics, said: “KatanaConnect enables the fastest, most future-proof path to feature-rich, connected AI...

IBM builds new AI chip for training deep learning models

IBM has created an upgraded version of its 2021 AI Telum processor to accelerate the development of industrial-scale deep learning models.

Called the Artificial Intelligence Unit (AIU), it is an application-specific integrated circuit that can run any deep learning task, from rapidly processing images to translating spoken languages.

Taking five years to research and develop, the AIU is computational processing hardware that is designed to handle AI calculations...

Edjx partners with Zeblok to boost AI app deployment at the edge

AI deployment

Decentralised edge computing provider Edjx has partnered with Zeblok Computational, an AI platform-as-a-service (PaaS) provider.

Zeblok’s AI-MicroCloud will integrate with the Edjx platform to meet the growing demand for simplified and rapid edge deployment solutions.

Edjx and Zeblok clients will benefit from this integration by having easier access to AI models for deployment across thousands of edge locations.

The AI-MicroCloud offered by Zeblok supports...

Stanford engineers build more efficient AI chip for edge devices

Stanford University engineers have created a more efficient and flexible AI chip well suited to power AI in tiny edge devices.

The engineers’ chip, called NeuRRAM, is a novel resistive random-access memory (RRAM) chip that innovates on the way current chips process and store data.

AI functionality on tiny edge devices today is limited by the energy a battery can provide. Whilst modern AI chips process and store data in separate locations – a compute unit and a memory...

MemryX begins customer sampling of new MX3 AI Accelerator

AI chip

Semiconductor developer MemryX, which designs AI processing solutions for edge devices, has begun customer sampling of its MX3 AI Accelerator.

The MX3 was created with ease-of-use and efficiency in mind. Able to achieve optimised model performance within minutes, MemryX hopes its accelerator will save companies months of software development time.

Built for edge computing systems old and new, the MX3 can accelerate AI processing in most systems and connects directly to...

Syntiant raises £42m to scale its global edge AI chip provision

Syntiant, a leading edge AI chip developer, has closed $55 million (£42m) in a series D funding round, bringing total investment in the company to $120 million (£91m) since its founding in 2017.

The Silicon Valley-based firm bagged five new investors including Renesas Electronics, Millennium Technology Value Partners, and Mirae Asset Capital, amongst others.

“We are at a pivotal point of our company’s growth and development, having shipped more than 20 million of...

Quadric pulls in £16m for on-device edge AI processing platform

Quadric, developers of an edge processing platform optimised for on-device AI at the network edge, have raised $21 million (£16m) in a Series B funding round.

Led by NSITEXE, the round also saw major new investment from MegaChips alongside existing investors Leawood VC, Pear VC, Uncork Capital, and Cota Capital.

Quadric’s platform is capable of handling neural backbones alongside classical dynamic data-parallel algorithms in a unified architecture. Whilst most other...

BrainChip brings neuromorphic edge AI chips to market

BrainChip Akida AKD1000 Processor

Edge AI firm BrainChip has brought its Akida AKD1000 processor to commercial markets following its usage in a new Mercedes-Benz concept electric car in January.

The Mercedes Vision EQXX, promoted as “the most efficient Mercedes-Benz ever built”, uses the Akida chip for in-cabin keyword spotting.

BrainChip’s products are specialised to learn like the human brain through neuromorphic AI architecture, a niche that helped the company grow in recognition last...

Deeplite platform accelerates AI on Arm CPUs through quantisation

CPU

Deeplite, a Canadian AI software company, has unveiled its Deeplite Runtime (DeepliteRT) platform that makes AI models smaller and faster without compromising accuracy.

Thanks to quantisation, customers will benefit from reduced power consumption and the ability to use existing Arm CPUs to run AI models with GPU-level performance.

AI optimisation’s importance is growing in an industry where edge devices are increasingly being turned to. Making AI models run on small...

NVIDIA unveils 200 trillion operations per second edge AI accelerator

NVIDIA

NVIDIA debuted the latest version in its AI accelerator series for robotics and edge devices at its GTC event on November 9, the Jetson AGX Orin.

The new model is built on NVIDIA’s Ampere architecture and can process up to 200 trillion operations per second, six times the capacity of its predecessor.

The company's Jetson series of products are used in a diverse range of edge and AI devices for a wide variety of use cases.

NVIDIA’s VP and general manager of...