Quantum Computing Report

Qrypt Partnering with NVIDIA to Provide Quantum Secure Cryptography Between AI Clusters

Qrypt, a quantum-secure encryption company, has announced the launch of its Quantum Entropy and Quantum Secure Key Generation technologies. These technologies will be used to transmit AI workloads using NVIDIA BlueField-3 DPUs with quantum-secure encryption, providing protection against quantum computing threats to current cryptography.

Most recently, Qrypt announced a Phase 1 contract with the Air Force to enhance national defense capabilities against quantum threats. 

The announcement of the partnership between NVIDIA and Qrypt is a logical next step in this ambition, as Qrypt provides highly secure entropy, specifically in this context:

Qrypt’s technology integrates directly into IPsec, which is traditionally used by NVIDIA BlueField DPU to set up encrypted connections between AI clusters. This integration allows for key generation and establishment without transmitting keys, establishing a quantum-secure channel. Qrypt’s solution uses true quantum random entropy to generate all key materials, ensuring the highest-quality encryption keys for securing the AI clusters. These keys are never transmitted across the network, eliminating the “harvest now, decrypt later” vulnerability. Qrypt’s Quantum Secure Encryption generates one-time pads and symmetric keys at multiple endpoints, providing unbreakable encryption.

The integration of Qrypt’s technology with the DPUs enables both North-South and East-West quantum secure traffic, with direct connections or over the internet. For example, if one DPU is in an enterprise data center and the other in an Nvidia data center, they can establish a quantum secure link even if the adversary controls all the external channels. Inside the data center, Qrypt has already shown this on multiple connected DPUs in different servers – this eliminates the risks from a nefarious sys admin or anyone in the data center with physical access to the DPUs from being able to harvest any useful data for exfiltration. Both issues have been a barrier for compliance industry companies using GPU clusters in the cloud to train AI data.

While all of these technologies remain to survive battle testing, they offer innovative approaches to new problems:

What we expect to see more of to assess current impact and future potential includes some of the following:

Denis Mandich, CTO of Qrypt, said “Our vision is to deploy this everywhere in the world and transform the cryptographic architecture of the internet itself, which was really designed and built around monetizing and data mining. We think that should go away and we should restore privacy the same way we had it before the internet existed. This is a way to do it.”

Qrypt is a member of NVIDIA Inception, a program designed to help startups evolve faster by providing support in areas including marketing, technology, and financing. You can view a press release with additional details about this development on the Qrypt website here.

March 15, 2024

Exit mobile version