Kinara
Click to visit website
About
Kinara provides high-performance, power-efficient Discrete Neural Processing Units (DNPUs) designed to bring advanced intelligence directly to edge devices. Now part of NXP Semiconductors, Kinara's technology enables local, responsive, and secure on-device decision-making for a variety of applications, from factory robots to smart cameras. By processing complex inferencing workloads locally, these chips eliminate the need for cloud dependency, thereby reducing latency, improving data privacy, and lowering bandwidth costs.\n\nThe platform revolves around the Ara series of processors, most notably the Ara-1 and Ara-2. The Ara-1 is a vision-centric first-generation chip delivering up to 6 eTOPS, while the Ara-2 is a more powerful second-generation unit capable of up to 40 eTOPS. Both chips utilize a programmable, RISC-V-based dataflow architecture that allows for efficient execution across proprietary multiply-accumulate (MAC) units. This design is specifically optimized for modern AI demands, including convolutional neural networks (CNNs), transformer-based architectures, and generative AI models like LLMs and VLMs.\n\nThese solutions are ideal for developers and engineers working in industries such as manufacturing, retail, and smart home technology. In practice, the DNPUs are often paired with NXP’s i.MX applications processors. In this synergy, the DNPU handles the heavy lifting of AI inferencing while the MPU manages preprocessing, postprocessing, and general-purpose compute tasks. This modular approach allows customers to scale AI performance independently of the main processor, providing the flexibility needed to meet specific cost or power budgets.\n\nWhat distinguishes Kinara is its comprehensive software ecosystem. The Kinara SDK, which includes model optimization tools and a library of pre-optimized AI models, is integrated into NXP’s eIQ software development environment. This integration offers a unified tooling experience, allowing developers to build and deploy production-ready solutions across a wide portfolio of processors and connectivity options. Whether implementing industrial safety analysis or real-time voice assistants, the combination of discrete hardware and robust software minimizes system complexity.
Pros & Cons
Delivers up to 40 eTOPS of performance for edge generative AI workloads.
Features a programmable RISC-V based dataflow design for algorithm flexibility.
Seamlessly integrates with NXP eIQ development environment for streamlined deployment.
Reduces latency and bandwidth costs by enabling local, cloud-independent inferencing.
Supports diverse architectures including both CNNs and transformer-based models.
Requires integration with application processors like i.MX for preprocessing tasks.
First-generation Ara-1 performance is significantly lower at only 6 eTOPS.
Use Cases
Industrial engineers can deploy Ara-2 units for real-time factory hazard detection and predictive maintenance without cloud reliance.
Smart camera developers can implement multimodal LLMs for context-aware security monitoring and elderly care.
Retail tech designers can utilize generative AI to create virtual showrooms tailored to individual customer preferences locally.
Robotics developers can use discrete NPUs to offload heavy vision tasks from general-purpose processors.
Platform
Features
• multimodal llm support
• cnn & transformer support
• 40 etops performance
• generative ai optimization
• eiq software integration
• risc-v dataflow architecture
• ara-2 dnpu
• ara-1 dnpu
FAQs
What is the peak performance of the Ara-2 DNPU?
The Ara-2 is a high-performance second-generation chip that delivers up to 40 eTOPS (equivalent TOPS). It is specifically optimized for demanding workloads such as generative AI, large language models (LLMs), and system-level acceleration.
Can these processors handle transformer-based AI models?
Yes, the Ara-2 architecture is built to handle both convolutional neural networks (CNNs) and modern transformer-based architectures. This makes it suitable for multimodal applications that combine visual inputs with text or speech for context-aware inferencing.
How does Kinara software integrate with NXP products?
The Kinara SDK is integrated into NXP’s eIQ software development environment. This allows developers to use a unified set of tools to build, optimize, and deploy AI applications across NXP’s combined portfolio of processors and NPUs.
What are the benefits of using a discrete NPU over cloud AI?
Using a discrete NPU like those from Kinara enables local inferencing, which significantly reduces latency and bandwidth costs. It also enhances data privacy and system resiliency by ensuring that decision-making happens on the device without needing a cloud connection.
Job Opportunities
Web Content Program Manager
Accelerate edge AI with high-performance discrete neural processing units that enable real-time generative AI and low-latency inferencing for hardware engineers.
Experience Requirements:
3–5 years of experience in web content operations, project management, or digital project coordination
Experience coordinating across IT, regional teams, and vendors
Working knowledge of CMS/web publishing platforms and translation tooling/workflows
Other Requirements:
English: B2 (Upper‑Intermediate) or higher
Spanish: Professional working proficiency
Responsibilities:
Optimize request intake processes
Deliver monthly reporting and performance summaries
Maintain governance frameworks and high‑quality documentation
Lead and execute content‑related projects as primary contact
Oversee the broader ecosystem of content projects for strategic alignment
Show more details
Technical Sales & Applications Program (TSAP)
Accelerate edge AI with high-performance discrete neural processing units that enable real-time generative AI and low-latency inferencing for hardware engineers.
Education Requirements:
Bachelor’s or Master’s degree in Electrical Engineering, Computer Engineering, or related field
Experience Requirements:
0-2 years of experience
Minimum GPA of 3.0 or higher
Other Requirements:
Willingness to relocate within the United States
Competitive mindset with a passion to win
Responsibilities:
Support customers and partners through evaluation, design, and deployment of NXP solutions
Translate customer requirements into system-level technical solutions
Collaborate with sales, applications, and product teams to drive design wins
Prepare technical presentations and demos
Lead customer engagements and internal reviews
Show more details
Netwerk en Control Engineer
Accelerate edge AI with high-performance discrete neural processing units that enable real-time generative AI and low-latency inferencing for hardware engineers.
Benefits:
Bonusplan
Mogelijkheid om extra vakantiedagen bij te kopen
Prima pensioenplan
Goed thuiswerkbeleid
Mogelijkheid om bedrijfsaandelen te kopen met 15% korting
Education Requirements:
HBO-nivo, door kennis en/of ervaring
Experience Requirements:
Inzicht van netwerken en kennis van scada systemen
Diepgaande (programmeer) kennis van PLC’s
Grondig inzicht in de functie van PLC’s in facilitaire installaties
Ruime ervaring met de meet- regelaspecten van facilitaire installaties
Other Requirements:
Bereidheid om mee te draaien in een meet- en regel consignatiedienst
Kosten bewust
Responsibilities:
Beheren van de procesbesturingen (M&R kasten) t.b.v. alle Facilitaire installaties
Beheer van Netwerk en Scada systemen (GBS)
Beheerder van hardware Security systemen
Participeren in projecten
Draagt zorg voor veiligheid, gezondheid, welzijn en milieu (VGWM)
Show more details
Ratings & Reviews
No ratings available yet. Be the first to rate this tool!
Featured Tools
adly.news
Connect with engaged niche audiences or monetize your subscriber base through an automated marketplace featuring verified metrics and secure Stripe payments.
View Details
Sceneform
Design hyper-realistic AI influencers and viral social media content with an all-in-one studio for persona building, motion syncing, and batch video rendering.
View Details
Grok Imagine
Transform creative ideas into cinematic 2K videos and photorealistic images with xAI’s Aurora engine, featuring precise motion control and multi-modal inputs.
View Details
Salespeak
Provide founder-level sales expertise across web, email, and LLM search with AI agents that learn your product in minutes to capture intent and convert buyers.
View Details
GPT Image 2
Transform text prompts and reference uploads into high-quality visuals with a streamlined browser-based generator designed for marketing and design workflows.
View Details
Seedance 2.0
Generate 2K cinematic videos with multi-shot storytelling and synchronized audio in under 60 seconds to transform text or images into professional-grade content.
View DetailsHappy Horse AI
Produce cinematic AI videos with native audio and consistent characters by combining text, images, and clips into beat-synced content for filmmakers and creators.
View Details
RemoveFrom.Video
Eliminate watermarks, subtitles, and unwanted objects from videos in seconds using AI-powered restoration that maintains high-quality footage and natural textures.
View Details