DEEPX and Wind River to advance mission-critical edge AI
Cette publication existe aussi en Français
DEEPX, an on-device AI semiconductor company, and Wind River have collaborated on a hardware and software combination for next-generation edge AI.
The partnership combines DEEPX’s on-device AI semiconductors with Wind River’s VxWorks real-time operating system (RTOS) and Wind River Helix Virtualisation Platform to help advance edge AI applications for mission-critical industries, including aerospace, defence, industrial, and robotics.
With the enablement of DEEPX’s neural processing units (NPUs) on Wind River technologies, companies can easily add accelerated AI to their real-time or edge systems at lower cost. DEEPX and Wind River collaborate to deliver a pre-validated stack that addresses safety and security concerns, enabling companies to reduce complexity and development cycles significantly.
“Edge AI is on a path to unlock incredible opportunities across a diverse range of industries. Together with Wind River’s long heritage in software for markets that are highly dependent on safety, security, reliability, and edge expertise, DEEPX can provide customers with innovative edge AI platforms that deliver exceptional performance, cost, and power efficiency,” said Lokwon Kim, CEO, DEEPX.
“The growth of real-time physical AI is creating immense potential for mission-critical environments while also presenting new, complex challenges. The collaboration between DEEPX’s advanced AI semiconductors and Wind River’s proven edge technologies will help advance AI to deliver tremendous outcomes across applications and industries,” said Avijit Sinha, Senior Vice President of Strategy and Global Business Development at Wind River.
DEEPX semiconductors optimise performance, reduce power consumption, and enhance cost efficiency across various industries, including smart camera modules, smart mobility, smart factories, consumer electronics, smart cities, surveillance systems, and AI servers. For example, the DX-M1 AI accelerator achieves 25 TOPS (200 eTOPS) with a power consumption of under 5 W, significantly outperforming competitors in terms of performance-to-power ratio. Building on the DX-M1 AI accelerator, the DX-M1M integrates high-performance LPDDR memory directly onto the same die. This design delivers the same industry-leading performance while dramatically reducing the overall footprint.
Study uses VR to speed training workers using robots
On-device AI model breakthrough streamlines development
AI accelerator brings low-power Generative AI to Raspberry Pi 5
Literal Labs in funding for energy-efficient logic-based AI models
The DX-V3 is a powerful AI vision processor that integrates a CPU, ISP, DSP, video codec, and high-performance NPU, delivering 13 TOPS of AI power with industry-leading efficiency. Optimised for next-gen vision systems, DX-V3 processes imagery, video streams, and 3D point-cloud data with ease.
The DX-M2 is a specialised semiconductor explicitly designed for the efficient processing of Generative AI and language models with up to 20 billion parameters. Delivering impressive performance of 25-30 TPS (Tokens Per Second), this processor enables the deployment of sophisticated language models in a variety of edge and enterprise applications without sacrificing response time or accuracy.
The first commercial RTOS to support Open Container Initiative (OCI) compliant containers, VxWorks enables companies to deploy new software-defined capabilities rapidly. The Helix Platform is a hypervisor that enables engineering teams to consolidate multiple systems onto a single, high-performance embedded system. It streamlines safety certification, reduces project risk, and accelerates time-to-market.
https://deepx.ai
www.windriver.com
If you enjoyed this article, you will like the following ones: don't miss them by subscribing to :
