Nvidia’s Morpheus AI security framework will land in April

GTC Nvidia teased several updates to its Morpheus AI security framework at GTC this week, and also announced that it would make the application framework generally available in April.

In addition to releasing a pre-built version of Morpheus, Nvidia will also release the framework’s full source code on GitHub to allow developers to modify Morpheus and build security applications on top of the software.

Since the chip design released Morpheus through an early access program nine months ago, nearly 700 developers and security vendors, including Cisco, F5, Lacework and Splunk, have created threat detection applications and log ingestion using Nvidia’s framework, Bartley Richardson, senior AI infrastructure manager at Nvidia, said during a security session on Tuesday.

And since it’s been a while since we last heard from Morpheus, Richardson also provided a quick refresher on the framework that Nvidia first started talking about last year. It’s “an AI cybersecurity framework designed to make inference in your security data streams easier, faster and more robust,” he said.

Specifically, Morpheus enables security developers to build AI pipelines that address specific use cases – such as fraud and phishing detection or sensitive information leak – by filtering and processing large volumes. data from logs and other network telemetry sources, including Nvidia BlueField DPUs. It is built on top of open-source QUICK software libraries, deep learning frameworks, and Nvidia’s Triton inference server.

“A lot has changed in Morpheus since our last update,” said Richardson. And those changes will be available when Morpheus moves from Early Access to General Availability next month.

Developer experience

Some of these changes are aimed at making it easier for developers to take advantage of GPUs for cybersecurity applications. To that end, the update will allow programmers to build pipelines from reusable steps in C++ or Python. It also adds support for multi-GPU execution without requiring the developer to write new code, allowing applications built on Morpheus to scale and process larger amounts of data.

Additionally, Nvidia has refined the API to allow for more customization and flexibility.

“We know performance matters when you’re analyzing traffic at bandwidth,” Richardson said. “So Morpheus now includes additional pipeline monitoring and inspection tools that allow you to capture precise performance metrics to verify that your pipelines are all working properly.”

Morpheus also got its own performance boost with faster speeds during data preprocessing and inference stages.

“During inference, we often want to classify items into buckets. These can be binary or multiclass classifications, and we’ve improved both binary classification for NLP workflows and binary classification for streams. FIL work, the first of more than 20 times and the scale of nearly 12 times,” said Richardson.

The new version of Morpheus can also extract raw anomaly scores from a model 200 times faster than the previous version, he noted. “This allows you to get a confidence score probability or an anomaly score of your model much faster,” Richardson explained. “And it allows you to act even sooner.”

Predefined fraud detection

In another new feature: a pre-built fraud detection use case will detect fraud out of the box using graphical neural networks to more accurately analyze more transactions and how those transactions interact with each other.

“First, node aggregation allows us to see how transactions from fraudulent nodes tend to connect abnormally with other fraudulent nodes,” Richardson said. “Second, malicious transactions are often linked to coordinated attacks. By observing these patterns, it becomes difficult for fraudsters to hide their behavior across the entire graph. There is nowhere to hide.”

In addition to crimes like credit card fraud, which Richardson predicts will cost the card industry more than $400 billion in fraud losses over the next decade, identity theft due to Fraud is also a growing threat to businesses and consumers.

“There were over a million reports of this in 2020, a 1,663% increase from just two years ago,” he noted. “Current methodologies are simply too slow, rely on pre-determined expert functionality, and require a substantial amount of labeled data to be effective. Next-generation fraud detection fills all of these gaps.” ®

Comments are closed.