AWS Admits Cloud Ain't Always The Answer, Intros On-prem Vid-analysing Box

Amazon Web Services, the outfit famous for pioneering pay-as-you-go cloud computing, has produced a bit of on-prem hardware that it will sell for a once-off fee.

The device is called the "AWS Panorama Appliance" and the cloud colossus describes it as a "computer vision (CV) appliance designed to be deployed on your network to analyze images provided by your on-premises cameras".

"AWS customers agree the cloud is the most convenient place to train computer vision models thanks to its virtually infinite access to storage and compute resources," states the AWS promo for the new box. But the post also admits that, for some, the cloud ain't the right place to do the job.

"There are a number of reasons for that: sometimes the facilities where the images are captured do not have enough bandwidth to send video feeds to the cloud, some use cases require very low latency," AWS's post states. Some users, it adds, "just want to keep their images on premises and not send them for analysis outside of their network".

Hence the introduction of the Panorama appliance, which is designed to ingest video from existing cameras and run machine learning models to do the classification, detection, and tracking of whatever your cameras capture.

Sometimes the facilities do not have enough bandwidth to send video feeds to the cloud

AWS imagines those ML models could well have been created in its cloud with SageMaker, and will charge you for cloud storage of the models if that's the case. The devices can otherwise run without touching the AWS cloud, although there is a charge of $8.33 per month per camera stream.

The appliance itself costs $4,000 up front.

Charging for hardware is not AWS's usual modus operandi. Its Outposts on-prem clouds are priced on a consumption model. The Snow range of on-prem storage and compute appliances are also rented rather than sold.

The Panorama appliance's specs page states that it contains Nvidia's Jetson Xavier AGX AI edge box, with 32GB RAM. The spec doesn't mention local storage, but lists a pair of gigabit ethernet ports, the same number of HDMI 2.0 slots, and two USB ports.

AWS announced the appliance at its re:invent gabfest in December 2020, when The Register opined that the cloudy concern may be taking a rare step into on-prem hardware, but by doing so would be eating the lunches of server-makers and video hardware specialists alike. Panorama turns out to not have quite the power to drive cloud services consumption as other Amazonian efforts, since the ML models it requires could come from SageMaker or other sources. That fact, and the very pre-cloud pricing scheme, mean the device could therefore be something of a watershed for AWS. ®

RECENT NEWS

Google Leverages AI To Automatically Lock Phones During Theft

Amid increasing incidents of mobile phone thefts, Google has launched an AI-based feature that automatically locks the s... Read more

Microsofts Emissions Surge Nearly 30% Amid AI Demand Growth

Microsoft has reported a nearly 30% increase in its emissions from 2020 to 2023, underscoring the challenges the tech gi... Read more

Impact Of AWS Leadership Change On The Global AI Race

The recent leadership transition at Amazon Web Services (AWS), with Adam Selipsky stepping down and Matt Garman taking t... Read more

The Global Impact Of App Stores On Technology And Economy

Since Apple launched its App Store in 2008, app stores have become a central feature of the digital landscape, reshaping... Read more

Alibaba's Cloud Investment Strategy: Fuelling AI Innovation And Growth

Alibaba Group's cloud business, Alibaba Cloud, has emerged as a powerhouse in the tech industry, spearheading innovation... Read more

Elon Musk Takes On Government 'Censorship': A Clash Of Titans In The Digital Arena

Elon Musk's recent endeavors to challenge government-led content takedowns mark a significant development in the ongoing... Read more