We are excited to bring Transform 2022 back in-person July 19 and virtually July 20 - 28. Join AI and data leaders for insightful talks and exciting networking opportunities. Register today!
During its March 2022 GPU Technology Conference (GTC) this week, Nvidia unveiled Isaac Nova Orin, a computing and sensor architecture powered by the company’s Jetson AGX Orin hardware. Nvidia says that Isaac Nova Orin comes with “all the compute and sensor hardware needed to design, build, and test autonomy” in autonomous mobile robots (AMRs) — types of robots that can understand and move through their environment without being overseen directly by an operator.
Warehousing and logistics organizations among others apply AMRs to tasks that’d be harmful to — or not possible for — teams of human workers. Using AI, compute, and a sophisticated set of sensors, AMRs can carry heavy loads while dynamically assessing and responding to their surroundings — assisting with tasks including locating, picking, and moving inventory.
An IDC survey found that over 70% of order fulfillment operations and warehouses that deploy AMRs have experienced double-digit improvement in KPIs like cycle time, productivity, and inventory efficiency. (Cycle time refers to the amount of time a team spends actually working on producing an item until the item is ready for shipment.) That’s perhaps why the global AMR market was worth roughly $1.67 million in 2020, according to Fortune Business Insights, and projected to growth to $8.7 billion by 2028.
Isaac Nova Orin and Jetson AGX Orin
Isaac Nova Orin, which will be available later this year, pairs two Jetson AGX Orin units to deliver up to 550 TOPS of power. In hardware, TOPS — which stands for “trillions of operations per second” — indicates how many computing operations, or basic math problems, a chip can handle over a short period of time. As my former colleague Jeremy Horowitz notes, TOPS, while often touted in marketing materials, aren’t necessarily the best way to measure a chip’s capabilities. But Nvidia is spotlighting Isaac Nova Orin’s other features, like its ability to process data in real time from up to six cameras, three lidars, and eight ultrasonic sensors from an AMR.
Over-the-air software management support is “preintegrated” in Isaac Nova Orin and the hardware is “calibrated and tested to work out of the box,” Nvidia says. Isaac Nova Orin includes tools necessary to simulate the robot as well as software modules designed to accelerate perception and navigation tasks and map different robots’ environments.
Alongside Isaac Nova Orin, Nvidia announced that the Jetson AGX Orin developer kit, which the company first detailed in November, is now available to customers for purchase. Readers will recall that Jetson AGX Orin delivers 275 TOPS of compute power and features Nvidia’s Ampere architecture GPU, Arm Cortex-A78AE CPUs, AI and vision accelerators, and high-speed chip-to-chip interfaces.
Microsoft, John Deere, Amazon, Hyundai, and JD.com are among the early adopters of Jetson AGX Orin. Developer kits start at $1,999, and production modules will be available in Q4 2022 for $399.
“As AI transforms manufacturing, healthcare, retail, transportation, smart cities and other essential sectors of the economy, demand for processing continues to surge,” Deepu Talla, VP of embedded and edge computing at Nvidia, said in a press release. “A million developers and more than 6,000 companies have already turned to Jetson. The availability of Jetson AGX Orin will supercharge the efforts of the entire industry as it builds the next generation of robotics and edge AI products.”
With Isaac Nova Orin and Jetson AGX Orin, Nvidia is competing for a slice of the rapidly growing edge computing segment. Generally speaking, “edge computing” encompasses computing and storage resources at the location where data is produced, including on — or near — AMRs. STL Partners recently estimated that the edge computing addressable market will grow from $10 billion in size in 2020 to $543 billion in 2030.
Edge computing offers several advantages compared with cloud-based technologies, but it isn’t without challenges. Keeping data locally means more locations to protect, with increased physical access allowing for different kinds of cyberattacks. (Some experts argue the decentralized nature of edge computing leads to increased security.) And compute is limited at the edge, which restricts the number of tasks that can be performed.
Even so, Gartner predicts that more than 50% of large organizations will deploy at least one edge computing application to support the internet of things or immersive experiences by the end of 2021, up from less than 5% in 2019. The number of edge computing use cases could jump even further in the upcoming years, with the firm expecting that more than half of large enterprises will have at least six edge computing use cases deployed by the end of 2023.
VentureBeat's mission is to be a digital town square for technical decision-makers to gain knowledge about transformative enterprise technology and transact. Learn more about membership.