Ambarella Expands N1 Edge GenAI Family With SoC Targeted at On-Premise Multi-Channel VLM and NN Processing in Under 20 Watts
Ambarella (NASDAQ: AMBA) has unveiled the N1-655 edge GenAI system-on-chip (SoC) during CES, expanding its N1 edge GenAI family. The new chip can decode 12 simultaneous 1080p30 video streams while processing video and running multiple vision-language models (VLMs) and CNNs, all under 20 watts of power—10-100x lower than cloud processors.
The N1-655 is designed for on-premise AI boxes, autonomous mobile robots (AMRs), and smart-city security applications. It supports popular models like Phi, Gemma, LLaVA-OneVision, and Llama without requiring internet connectivity. To support N1-655 based designs, Ambarella is launching the Cooper Pro developer kit, which will begin shipping after CES.
This development addresses the growing trend toward edge GenAI processing, offering advantages over cloud-based solutions in terms of bandwidth costs, service fees, power usage, and data security. The N1-655 samples are currently available upon request.
Ambarella (NASDAQ: AMBA) ha presentato il N1-655 edge GenAI system-on-chip (SoC) durante il CES, ampliando la sua famiglia di chip N1 edge GenAI. Il nuovo chip è in grado di decodificare 12 flussi video simultanei a 1080p30 mentre elabora video e gestisce più modelli di visione-linguaggio (VLM) e reti neurali convoluzionali (CNN), tutto sotto 20 watt di potenza—da 10 a 100 volte meno rispetto ai processori cloud.
Il N1-655 è progettato per applicazioni di AI on-premise, robot mobili autonomi (AMR) e sicurezza nelle smart city. Supporta modelli popolari come Phi, Gemma, LLaVA-OneVision e Llama senza necessità di connettività internet. Per supportare i progetti basati su N1-655, Ambarella lancerà il kit sviluppatore Cooper Pro, che inizierà a essere spedito dopo il CES.
Questo sviluppo risponde alla crescente tendenza verso l'elaborazione GenAI edge, offrendo vantaggi rispetto alle soluzioni basate su cloud in termini di costi di larghezza di banda, costi di servizio, consumo energetico e sicurezza dei dati. I campioni del N1-655 sono attualmente disponibili su richiesta.
Ambarella (NASDAQ: AMBA) ha presentado el N1-655 edge GenAI system-on-chip (SoC) durante el CES, ampliando su familia de chips N1 edge GenAI. El nuevo chip puede decodificar 12 flujos de video simultáneos a 1080p30 mientras procesa video y ejecuta múltiples modelos de visión-lenguaje (VLM) y CNN, todo con menos de 20 vatios de potencia—de 10 a 100 veces menos que los procesadores en la nube.
El N1-655 está diseñado para aplicaciones de AI en las instalaciones, robots móviles autónomos (AMR) y seguridad en ciudades inteligentes. Admite modelos populares como Phi, Gemma, LLaVA-OneVision y Llama sin requerir conectividad a internet. Para apoyar los diseños basados en N1-655, Ambarella lanzará el kit de desarrollo Cooper Pro, que comenzará a enviarse después del CES.
Este desarrollo responde a la creciente tendencia hacia el procesamiento GenAI en el borde, ofreciendo ventajas sobre las soluciones basadas en la nube en términos de costos de ancho de banda, tarifas de servicio, consumo de energía y seguridad de datos. Las muestras del N1-655 están actualmente disponibles bajo solicitud.
앰바렐라 (NASDAQ: AMBA)는 CES에서 N1-655 엣지 GenAI 시스템 온 칩 (SoC)을 공개하며 N1 엣지 GenAI 가족을 확장했습니다. 새로운 칩은 12개의 동시 1080p30 비디오 스트림을 디코딩할 수 있으며, 비디오를 처리하고 다양한 비전-언어 모델 (VLM) 및 CNN을 실행할 수 있으며, 모두 20와트 이하의 전력으로 — 클라우드 프로세서보다 10-100배 낮습니다.
N1-655는 온프레미스 AI 박스, 자율 이동 로봇 (AMR) 및 스마트 시티 보안 애플리케이션을 위해 설계되었습니다. 인터넷 연결 없이도 Phi, Gemma, LLaVA-OneVision 및 Llama와 같은 인기 모델을 지원합니다. N1-655 기반 디자인을 지원하기 위해 앰바렐라는 CES 이후 배송을 시작할 Cooper Pro 개발자 키트를 출시할 예정입니다.
이 개발은 엣지 GenAI 처리로의 증가하는 추세에 대응하며, 대역폭 비용, 서비스 요금, 전력 사용 및 데이터 보안 측면에서 클라우드 기반 솔루션에 비해 이점을 제공합니다. N1-655 샘플은 현재 요청 시 이용 가능합니다.
Ambarella (NASDAQ: AMBA) a dévoilé le N1-655 edge GenAI system-on-chip (SoC) lors du CES, élargissant sa famille de puces N1 edge GenAI. La nouvelle puce peut décoder 12 flux vidéo simultanés à 1080p30 tout en traitant des vidéos et en exécutant plusieurs modèles de vision-langage (VLM) et des CNN, le tout avec moins de 20 watts de puissance—10 à 100 fois moins que les processeurs cloud.
Le N1-655 est conçu pour des applications d'IA sur site, de robots mobiles autonomes (AMR) et de sécurité dans les villes intelligentes. Il prend en charge des modèles populaires tels que Phi, Gemma, LLaVA-OneVision et Llama sans nécessiter de connexion internet. Pour soutenir les conceptions basées sur le N1-655, Ambarella lancera le kit de développement Cooper Pro, qui commencera à être expédié après le CES.
Ce développement répond à la tendance croissante vers le traitement GenAI en périphérie, offrant des avantages par rapport aux solutions basées sur le cloud en termes de coûts de bande passante, de frais de service, de consommation d'énergie et de sécurité des données. Des échantillons du N1-655 sont actuellement disponibles sur demande.
Ambarella (NASDAQ: AMBA) hat während der CES das N1-655 Edge GenAI System-on-Chip (SoC) vorgestellt und damit seine N1 Edge GenAI Familie erweitert. Der neue Chip kann 12 gleichzeitige 1080p30 Videostreams decodieren, während er Videos verarbeitet und mehrere vision-sprachliche Modelle (VLM) sowie CNNs ausführt, alles unter 20 Watt Leistung—10-100 Mal niedriger als Cloud-Prozessoren.
Das N1-655 ist für On-Premise AI-Boxen, autonome mobile Roboter (AMR) und Sicherheitsanwendungen in Smart Cities konzipiert. Es unterstützt gängige Modelle wie Phi, Gemma, LLaVA-OneVision und Llama, ohne Internetverbindung zu benötigen. Um Designs basierend auf N1-655 zu unterstützen, wird Ambarella das Entwicklerkit Cooper Pro auf den Markt bringen, das nach der CES versendet wird.
Diese Entwicklung reagiert auf den wachsenden Trend zur Edge GenAI-Verarbeitung und bietet Vorteile gegenüber Cloud-basierten Lösungen in Bezug auf Bandbreitenkosten, Servicegebühren, Energieverbrauch und Datensicherheit. Die Muster des N1-655 sind derzeit auf Anfrage verfügbar.
- New N1-655 SoC achieves 10-100x lower power consumption than cloud processors
- Capability to process 12 simultaneous 1080p30 video streams
- Supports multiple popular AI models without internet connectivity requirement
- Addresses growing edge computing market while reducing operational costs
- None.
Insights
The N1-655 SoC represents a significant technological leap in edge AI processing capabilities. The ability to decode 12 simultaneous 1080p30 video streams while running multiple VLMs and CNNs under 20W power consumption is remarkable, offering 10-100x lower power consumption than cloud processors. The support for popular models like Phi, Gemma, LLaVA-OneVision and Llama positions this chip favorably in the growing edge AI market.
In simple terms, imagine running sophisticated AI that can understand videos, images and text simultaneously on a device that uses about as much power as a bright LED bulb, instead of requiring a power-hungry server room. This efficiency could dramatically reduce operational costs for businesses deploying AI solutions.
The N1-655 strategically targets three growing markets: on-premise AI boxes, autonomous mobile robots (AMRs) and smart city security. This positions Ambarella to capitalize on the increasing demand for edge AI solutions driven by privacy concerns and cloud computing costs. The edge AI semiconductor market is experiencing rapid growth and this product addresses key pain points around power consumption, cost and data privacy.
For investors, this represents Ambarella's strong execution in expanding their edge AI portfolio and potential market share growth in the robotics and smart city sectors. The timing of this release during CES 2025, coupled with immediate sample availability, indicates confidence in the product's market readiness.
This product launch strengthens Ambarella's competitive position in the edge AI semiconductor space. The focus on power efficiency and local processing directly addresses the growing concerns about cloud computing costs, potentially opening up new revenue streams in industrial and commercial applications. The expansion of the Cooper Developer Platform with Cooper Pro demonstrates a comprehensive ecosystem approach, which could accelerate adoption and create recurring revenue opportunities.
From a business perspective, this positions Ambarella to capture value in both hardware sales and potential software/development tool revenues, while the immediate availability of samples suggests potential revenue impact in the near term.
New AI SoC is Ideal for On-Premise AI Boxes, Autonomous Robotics and Smart City Security; Brings High-Performance GenAI to Power- and Cost-Constrained Edge AI Applications
SANTA CLARA, Calif., Jan. 07, 2025 (GLOBE NEWSWIRE) -- Ambarella, Inc. (NASDAQ: AMBA), an edge AI semiconductor company, today announced during CES the N1-655 edge GenAI system-on-chip (SoC), which provides on-chip decode of 12x simultaneous 1080p30 video streams, while concurrently processing that video and running a hybrid of multiple, multimodal vision-language models (VLMs) and traditional CNNs. This SoC’s high AI processing performance supports most of the popular multimodal VLMs and large-language models (LLMs), while consuming only 20 watts of power—10-100x lower than cloud processors. For example, the N1-655 reliably runs the Phi, Gemma, LLaVA-OneVision and Llama models, without the need for an internet connection, on data inputs like visuals and speech in applications such as on-premise AI boxes, autonomous mobile robots (AMRs), and smart-city security video recorders.
Following its initial N1 SoC introduced last year, Ambarella is building up a family of edge GenAI SoCs for tasks that go beyond what can be done on-camera. Both of the current family members process GenAI models locally to improve privacy with state-of-the-art performance per watt, significantly reducing power consumption as well as the total cost of ownership compared to cloud-based inference processing.
“The N1-655 continues Ambarella’s industry-leading levels of AI performance per watt, this time for AMRs and on-premise AI box applications that require the latest edge GenAI features while keeping power consumption under 20 watts,” said Fermi Wang, President and CEO of Ambarella. “We tailored this new SoC to run faster, while fine-tuning the most popular LLMs and VLMs to make them even smaller, eliminating the need for high-bandwidth and cost-intensive cloud processing in this class of applications.”
There is a trend toward doing more GenAI processing at the edge, due to cloud-based processing’s bandwidth cost, ongoing service fees and high power usage, along with data security and privacy concerns. Ambarella employs cutting-edge fine tuning techniques such as QLoRA, which reduces the footprint sizes of popular GenAI models to enable their use in edge applications. The Company is at the forefront of this trend, as exemplified by the high performance and low power consumption of the new N1-655.
To support N1-655 based designs, Ambarella is also expanding its Cooper™ Developer Platform with the new Cooper Pro member of the Cooper developer kit family, which provides industrial-class performance. This kit is powered by the N1-655 and will begin shipping after CES. To register interest in the Cooper Pro, visit www.ambarella.com/cooper.
N1-655 samples are available now, upon request. This new SoC and the Cooper Pro kit are also being demonstrated at Ambarella’s invitation-only exhibition during CES in Las Vegas this week. For more information or to schedule a demo during the show, please contact your Ambarella representative or visit www.ambarella.com/products/aiot-industrial-robotics.
About Ambarella
Ambarella’s products are used in a wide variety of human vision and edge AI applications, including video security, advanced driver assistance systems (ADAS), electronic mirror, drive recorder, driver/cabin monitoring, autonomous driving and robotics applications. Ambarella’s low-power systems-on-chip (SoCs) offer high-resolution video compression, advanced image and radar processing, and powerful deep neural network processing to enable intelligent perception, fusion and planning. For more information, please visit www.ambarella.com.
Ambarella Contacts
- Media contact: Eric Lawson, elawson@ambarella.com, +1 480-276-9572
- Investor contact: Louis Gerhardy, lgerhardy@ambarella.com, +1 408-636-2310
- Sales contact: https://www.ambarella.com/contact-us/
All brand names, product names, or trademarks belong to their respective holders. Ambarella reserves the right to alter product and service offerings, specifications, and pricing at any time without notice. © 2025 Ambarella. All rights reserved.
A photo accompanying this announcement is available at https://www.globenewswire.com/NewsRoom/AttachmentNg/99505458-f5b9-4c1a-b744-ffa97b0c7bf5
FAQ
What are the key features of Ambarella's new N1-655 GenAI chip (AMBA)?
How much power does the Ambarella N1-655 (AMBA) consume compared to cloud processors?
What applications is the Ambarella N1-655 (AMBA) designed for?
When will Ambarella's Cooper Pro developer kit for N1-655 (AMBA) be available?
Are samples of Ambarella's N1-655 (AMBA) currently available?
