NVIDIA SDKs have been instrumental in accelerating AI applications across a spectrum of use cases spanning smart cities, medical, and robotics. However, achieving a production-grade AI solution that can deployed at the edge to support human and machine collaboration safely and securely needs both high-quality hardware and software tailored for enterprise needs.
NVIDIA is again accelerating mission-critical AI applications at the edge. NVIDIA AI Enterprise IGX is an enterprise-grade software solution for running edge AI on the NVIDIA IGX platforms.
NVIDIA IGX provides your industrial-grade hardware with ultra-fast AI performance and high throughput data processing in the size and power envelope needed for the edge, along with robust edge security, remote manageability, and built-in safety mechanisms.
The integration of NVIDIA IGX with NVIDIA AI Enterprise makes AI at the edge more accessible, efficient, and enterprise-friendly.
Edge AI requirements for enterprises
AI stacks are made of and rely on many third-party applications and open-source software (OSS). The complexity of maintaining the security and stability of an AI software stack with increasing dependencies is a massive undertaking. This resource-intensive task hinders the pace of AI innovation for enterprises.
Here are some of the factors to consider:
- Stability and reliability
- Security and threat mitigation
- Regulatory issues
Stability and reliability
Understanding all the dependencies within the software packages is critical to maintaining API stability. A change in any of the third-party and OSS packages may require a subsequent change in a dependent API, which can break the stack. This web of dependencies makes it nearly impossible for enterprises to standardize the latest open-source versions in their deployments.
Security and threat mitigation
The ever-expanding software dependencies increase the risk of introducing software vulnerabilities. These could lead to an attack with severe consequences, such as data breaches, disruption of services, and huge financial losses.
The number of critical vulnerabilities reported in 2022 is up 59% in comparison to 2021, totaling 4135 critical vulnerabilities. The industries with higher standards for latency, privacy, bandwidth, and regulatory requirements are not immune to this trend either. According to the Clinicians’ Perspectives on Healthcare Cybersecurity and Cyber Threats article, in 2019, 24% of cyberattacks were in the healthcare industry. Cyberattacks on hospitals surged in 2022.
Regulatory issues
Many highly regulated industries impose stringent regulations for AI software maintenance.
For example, in the aerospace sector, where AI is increasingly used in navigation systems and flight control algorithms, strict guidelines ensure continuous monitoring and updates to maintain system integrity and safety standards.
Similarly, in the healthcare industry, AI-driven therapeutic and diagnostic tools must adhere to rigorous maintenance protocols to guarantee accuracy and reliability in medical decision-making processes.
In 2023, the FDA issued the final guidance document, Cybersecurity in Medical Devices: Quality System Considerations and Content of Premarket Submissions, which states the Refuse To Accept (RTA) policy for premarket submissions that do not contain the information required by section 524B (“Ensuring Cybersecurity of Devices”). The FDA can now reject new medical devices through cybersecurity standards.
NVIDIA AI Enterprise IGX
NVIDIA AI Enterprise IGX is a purpose-built software solution for mission-critical applications that run on the NVIDIA IGX Orin platforms.
It provides unmatched performance, security, stability, and support for the entire software stack, relieving enterprises of the intricacies associated with maintaining and securing complex AI software platforms. This not only streamlines AI-powered operations but also instills a higher level of confidence in deploying AI applications at scale.
Building on the success of NVIDIA AI Enterprise for data centers and cloud, the introduction of NVIDIA AI Enterprise IGX extends these advantages to the edge.
As an industrial-grade, high-performance, enterprise-level edge solution, together with NVIDIA IGX, NVIDIA AI Enterprise IGX provides comprehensive, end-to-end software support across firmware, OS, drivers, AI frameworks, SDKs, and libraries. With a stable software distribution purpose-built for NVIDIA IGX, you can expect hardware and software support for the extended timelines required by critical use cases and rely on a single source of support, long-term API stability, and a consistent software bill of materials (SBOM).
The value of included enterprise support provides a solution for addressing technical issues promptly to ensure the stability and reliability of edge AI systems, avoiding downtime that can lead to significant financial losses or operational disruptions.
NVIDIA AI Enterprise IGX includes essential AI libraries and SDKs, such as NVIDIA TensorRT, TensorFlow, PyTorch, NVIDIA Triton Inference Server, Holoscan, and an enterprise Ubuntu Linux operating system.
There are two add-on options:
- Yocto: For enterprises deploying Yocto.
- Functional safety: Where safety, people, and property are paramount, as in industrial manufacturing. This ensures that the system can protect critical assets and people’s lives from both proactive and reactive safety approaches.
NVIDIA AI Enterprise IGX is designed to empower enterprises to build cutting-edge technology while reducing TCO, speeding time to market, and making product development more efficient overall.
It comes with flexible software branches to provide you with choices, depending on your industry needs and product development cycle:
- Production branches: Ensure API stability with monthly security updates. These branches are ideal for deploying AI in production when stability is required. Released every 6 months with a 9-month lifecycle.
- Long-term support branches: Support long-term application lifecycle with quarterly security updates. These branches are ideal for highly regulated AI applications. Released every 2.5 years with a lifecycle of up to 10 years.
The solution offers a proven path to AI success in the challenging and complex edge AI development process.
Get started with NVIDIA AI Enterprise IGX
If you are in the process of developing edge AI applications or are ready for production, get started with NVIDIA AI Enterprise today:
- Start developments with the NVIDIA IGX Orin Developer Kit
- Contact NVIDIA to request an NVIDIA AI Enterprise IGX software trial
- Learn more about NVIDIA AI Enterprise IGX
