Check Point is proud to integrate with NVIDIA DSX Air’s testing environment, enabling organizations to pre-validate their security aware AI data center designs before ever deploying their first piece of hardware in production to build and run their own AI.
Testing AI Factory deployments end-to-end is challenging and can require complex multi-vendor orchestration. From compute to networking, orchestration, and security, ensuring integrations, configurations and automations perform as expected can become resource-intensive with so many factors at play.
Now, organizations can perform large-scale cyber security validation testing before deploying AI Factories, using the NVIDIA DSX Air cloud-based simulation and validation platform.
Why are Organizations Building Their Own AI?
Considerations such as sovereign AI mandates, high cloud-compute costs, and risk to intellectual property are driving enterprises to develop their own AI infrastructure.
When building these proprietary AI data centers, or “AI factories”, organizations often opt to build new infrastructure from the ground up, or alternatively, transform their current data center into one that can develop and run AI applications.
By evolving traditional data centers into AI factories with private LLMs at their core, enterprises are looking to reimagine their businesses from the inside out, converting huge volumes of raw enterprise data into powerful business intelligence and productivity boosts.
The term “AI Factory” describes the production of high-value AI tokens and actionable analysis that can leverage raw data to suggest innovative new business models, define profitable customer offers, unleash productivity gains, and drive significant cost savings.
AI Data Centers – What Are the Risks?
Bringing AI into data centers requires rethinking the entire cyber security architecture; the new threat surface expands far beyond traditional IT. The integration of AI systems and agents combined with hyperconnectivity to vast enterprise databases and systems creates the ultimate target-rich environment for threat actors of all kinds, from nation state actors to everyday hackers.
Deploying AI systems in enterprise data centers (the most highly protected environment) also introduces an entirely new breed of risks including:
- System vulnerabilities across the AI infrastructure
- Dangerous east-west lateral movement within AI clusters
- AI supply chain failures and weakened system resilience
- LLM prompt injections or subtle manipulations
- Model theft, model abuse, and poisoning attacks
- Data loss/theft and PII leakage via illicit prompts, responses, or unprotected logs
- Inconsistent, fragmented security policies across environments including inference clusters, AI apps, and gateways
- Failure to meet mandatory government compliance and industry regulations
Validating AI Factory Security with Check Point and NVIDIA
With NVIDIA DSX Air, organizations can validate configurations, features, and automations, simulating full-scale network architectures that may be running various network operating systems.
Check Point’s AI Factory Firewall and AI security stack operate in the DSX Air environment, enabling full security validation across the NVIDIA infrastructure and software, along with third party ecosystem solutions from storage, networking, and Kubernetes vendors.
Check Point also enables enterprises to apply security policies for agents and users in a multi-tenant environment.
Lowering the Risk of AI Data Center Deployments
NVIDIA DSX Air lowers the risk and cost of data center migration to high performance AI factory solutions. It is designed for large-scale deployments (thousands of GPUs) including NVIDIA Spectrum-X Ethernet network switches, ConnectX SuperNICs, hosts, BlueField DPUs, NVLink and NVIDIA AI software stack. Enterprises can perform cyber security checks on a replica of their AI Factory production design (a “digital twin”) to detect security events and vulnerabilities before going live, to ensure strict security starting at token number one.
Check Point and NVIDIA – Collaborating to Secure your AI Infrastructure at Scale
From AI model development to agentic applications and inference, Check Point and NVIDIA are partnering to secure the entire AI lifecycle for enterprises and service providers, offering:
- Tamper-proof code, network, and host-level security
- Validation for everyday enterprise use with NVIDIA RTX PRO servers
- Future-ready technology running seamlessly on NVIDIA Bluefield-4 DPUs expected in H2 2026
See the Demo at NVIDIA GTC Event
At GTC 2026, users can see a demo of NVIDIA DSX Air with the Check Point AI Factory Firewall running on simulated BlueField DPUs. The firewall provides critical security including tenant isolation to secure the AI data center. Simulated AI traffic is generated by Keysight Cyperf software, which also runs in NVIDIA Air.
To secure your AI infrastructure, check out Check Point’s AI Factory Firewall , or explore NVIDIA’s safe simulation environment here: NVIDIA DSX Air.
Join the live session on DSX Air here.