November 24, 2021
Nvidia is fast-tracking its cybersecurity efforts, emphasizing zero trust through new product integrations designed to protect enterprise customers from attack while supporting artificial intelligence, machine learning and server workloads that scale. Earlier this month Nvidia promoted its full-stack data center security solution: DOCA 1.2 accelerated software, running on BlueField-3 DPUs using the Morpheus AI framework — a configuration that can “secure a data center at every touchpoint,” including users, devices and the data itself, Nvidia founder and CEO Jensen Huang explained at Nvidia’s GTC 2021 event earlier this month.
Container-based cloud solutions allow hyperscale computers the ability to nimbly process vast quantities of data, but such performance comes at a cost: opening data center ports that must be secured from attack. Protection at the perimeter (a firewall) and workgroup segmentation (which divides a network into subnets with different levels of access) “are no longer sufficient,” Huang said during his GTC 2021 keynote.
Basically an embedded security monitoring system, that runs through every level of a computer infrastructure, Morpheus can protect against dynamic threats in real time. As such it meets the zero trust parameters that President Biden established via executive order for federal entities, while also urging them on private business leaders.
“Top-down approaches to fixing legacy tech stacks to support zero trust are hard to do well,” says VentureBeat, explaining why server replacements along the lines of what Nvidia is proposing are going to be necessary, so don’t expect cost-cutting through mere upgrades.
You can run systems on Nvidia’s cloud-native Quantum-2, a 400GB-per-second InfiniBand platform Huang calls “the first networking platform to offer the performance of a super-computer and the shareability of cloud computing.” Combined with the $7 billion acquisition of high-performance networking firm Mellanox in 2020, Nvidia is loaded for bear.
“Designed to detect threats and breach attempts using unsupervised machine learning algorithms, Nvidia Morpheus is a continuously learning cybersecurity framework that adapts and modifies workflows based on accumulated data patterns,” VentureBeat explains.
“The rise of transformative innovations such as AI, 5G, and smart devices has dramatically expanded traffic moving through the modern data center, making it more difficult to identify potential breaches and attacks,” said Kevin Deierling, Nvidia senior vice president of networking.